minix/drivers/dp8390/dp8390.c

1873 lines
46 KiB
C
Raw Normal View History

/*
* dp8390.c
*
* This file contains a ethernet device driver for NS dp8390 based ethernet
* cards.
*
* Created: before Dec 28, 1992 by Philip Homburg <philip@f-mnx.phicoh.com>
*
* Modified Mar 10 1994 by Philip Homburg
* Become a generic dp8390 driver.
*
* Modified Dec 20 1996 by G. Falzoni <falzoni@marina.scn.de>
* Added support for 3c503 boards.
*/
#include <minix/drivers.h>
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
#include <minix/netdriver.h>
#include <stdlib.h>
#include <minix/com.h>
#include <minix/endpoint.h>
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
#include <minix/ds.h>
#include <net/hton.h>
#include <net/gen/ether.h>
#include <net/gen/eth_io.h>
#include <machine/vm.h>
#include <sys/mman.h>
#include "assert.h"
#include "local.h"
#include "dp8390.h"
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
static dpeth_t de_state;
static int de_instance;
u32_t system_hz;
/* Configuration */
typedef struct dp_conf
{
port_t dpc_port;
int dpc_irq;
phys_bytes dpc_mem;
} dp_conf_t;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
#define DP_CONF_NR 4
2012-03-25 20:25:53 +02:00
static dp_conf_t dp_conf[DP_CONF_NR]= /* Card addresses */
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
/* I/O port, IRQ, Buffer address. */
{ 0x280, 3, 0xD0000, },
{ 0x300, 5, 0xC8000, },
{ 0x380, 10, 0xD8000, },
{ 0x000, 0, 0x00000, },
};
/* Card inits configured out? */
#if !ENABLE_WDETH
#define wdeth_probe(dep) (0)
#endif
#if !ENABLE_NE2000
#define ne_probe(dep) (0)
#endif
#if !ENABLE_3C503
#define el2_probe(dep) (0)
#endif
/* Some clones of the dp8390 and the PC emulator 'Bochs' require the CR_STA
* on writes to the CR register. Additional CR_STAs do not appear to hurt
* genuine dp8390s
*/
#define CR_EXTRA CR_STA
#if ENABLE_PCI
static void pci_conf(void);
#endif
static void do_vwrite_s(message *mp, int from_int);
static void do_vread_s(message *mp);
static void do_init(message *mp);
static void do_int(dpeth_t *dep);
static void do_getstat_s(message *mp);
static void dp_stop(dpeth_t *dep);
static void dp_init(dpeth_t *dep);
static void dp_confaddr(dpeth_t *dep);
static void dp_reinit(dpeth_t *dep);
static void dp_reset(dpeth_t *dep);
static void dp_check_ints(dpeth_t *dep);
static void dp_recv(dpeth_t *dep);
static void dp_send(dpeth_t *dep);
static void dp_getblock(dpeth_t *dep, int page, size_t offset, size_t
size, void *dst);
static void dp_pio8_getblock(dpeth_t *dep, int page, size_t offset,
size_t size, void *dst);
static void dp_pio16_getblock(dpeth_t *dep, int page, size_t offset,
size_t size, void *dst);
static int dp_pkt2user_s(dpeth_t *dep, int page, vir_bytes length);
static void dp_user2nic_s(dpeth_t *dep, iovec_dat_s_t *iovp, vir_bytes
offset, int nic_addr, vir_bytes count);
static void dp_pio8_user2nic_s(dpeth_t *dep, iovec_dat_s_t *iovp,
vir_bytes offset, int nic_addr, vir_bytes count);
static void dp_pio16_user2nic_s(dpeth_t *dep, iovec_dat_s_t *iovp,
vir_bytes offset, int nic_addr, vir_bytes count);
static void dp_nic2user_s(dpeth_t *dep, int nic_addr, iovec_dat_s_t
*iovp, vir_bytes offset, vir_bytes count);
static void dp_pio8_nic2user_s(dpeth_t *dep, int nic_addr, iovec_dat_s_t
*iovp, vir_bytes offset, vir_bytes count);
static void dp_pio16_nic2user_s(dpeth_t *dep, int nic_addr,
iovec_dat_s_t *iovp, vir_bytes offset, vir_bytes count);
static void dp_next_iovec_s(iovec_dat_s_t *iovp);
static void conf_hw(dpeth_t *dep);
static void update_conf(dpeth_t *dep, dp_conf_t *dcp);
static void map_hw_buffer(dpeth_t *dep);
static int calc_iovec_size_s(iovec_dat_s_t *iovp);
static void reply(dpeth_t *dep);
static void mess_reply(message *req, message *reply);
static void get_userdata_s(int user_proc, cp_grant_id_t grant, vir_bytes
offset, vir_bytes count, void *loc_addr);
static void put_userdata_s(int user_proc, cp_grant_id_t grant, size_t
count, void *loc_addr);
static void insb(port_t port, void *buf, size_t size);
static void insw(port_t port, void *buf, size_t size);
static void do_vir_insb(port_t port, int proc, vir_bytes buf, size_t
size);
static void do_vir_insw(port_t port, int proc, vir_bytes buf, size_t
size);
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
/* SEF functions and variables. */
2012-03-25 20:25:53 +02:00
static void sef_local_startup(void);
static int sef_cb_init_fresh(int type, sef_init_info_t *info);
static void sef_cb_signal_handler(int signo);
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
2012-03-25 20:25:53 +02:00
static void handle_hw_intr(void)
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
int r, irq;
dpeth_t *dep;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep = &de_state;
if (dep->de_mode != DEM_ENABLED)
return;
assert(dep->de_flags & DEF_ENABLED);
irq= dep->de_irq;
assert(irq >= 0 && irq < NR_IRQ_VECTORS);
if (dep->de_int_pending || 1)
{
dep->de_int_pending= 0;
dp_check_ints(dep);
do_int(dep);
r= sys_irqenable(&dep->de_hook);
if (r != OK) {
panic("unable enable interrupts: %d", r);
}
}
}
/*===========================================================================*
* dpeth_task *
*===========================================================================*/
int main(int argc, char *argv[])
{
message m;
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
int ipc_status;
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
int r;
Basic System Event Framework (SEF) with ping and live update. SYSLIB CHANGES: - SEF must be used by every system process and is thereby part of the system library. - The framework provides a receive() interface (sef_receive) for system processes to automatically catch known system even messages and process them. - SEF provides a default behavior for each type of system event, but allows system processes to register callbacks to override the default behavior. - Custom (local to the process) or predefined (provided by SEF) callback implementations can be registered to SEF. - SEF currently includes support for 2 types of system events: 1. SEF Ping. The event occurs every time RS sends a ping to figure out whether a system process is still alive. The default callback implementation provided by SEF is to notify RS back to let it know the process is alive and kicking. 2. SEF Live update. The event occurs every time RS sends a prepare to update message to let a system process know an update is available and to prepare for it. The live update support is very basic for now. SEF only deals with verifying if the prepare state can be supported by the process, dumping the state for debugging purposes, and providing an event-driven programming model to the process to react to state changes check-in when ready to update. - SEF should be extended in the future to integrate support for more types of system events. Ideally, all the cross-cutting concerns should be integrated into SEF to avoid duplicating code and ease extensibility. Examples include: * PM notify messages primarily used at shutdown. * SYSTEM notify messages primarily used for signals. * CLOCK notify messages used for system alarms. * Debug messages. IS could still be in charge of fkey handling but would forward the debug message to the target process (e.g. PM, if the user requested debug information about PM). SEF would then catch the message and do nothing unless the process has registered an appropriate callback to deal with the event. This simplifies the programming model to print debug information, avoids duplicating code, and reduces the effort to print debug information. SYSTEM PROCESSES CHANGES: - Every system process registers SEF callbacks it needs to override the default system behavior and calls sef_startup() right after being started. - sef_startup() does almost nothing now, but will be extended in the future to support callbacks of its own to let RS control and synchronize with every system process at initialization time. - Every system process calls sef_receive() now rather than receive() directly, to let SEF handle predefined system events. RS CHANGES: - RS supports a basic single-component live update protocol now, as follows: * When an update command is issued (via "service update *"), RS notifies the target system process to prepare for a specific update state. * If the process doesn't respond back in time, the update is aborted. * When the process responds back, RS kills it and marks it for refreshing. * The process is then automatically restarted as for a buggy process and can start running again. * Live update is currently prototyped as a controlled failure.
2009-12-21 15:12:21 +01:00
/* SEF local startup. */
env_setargs(argc, argv);
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
sef_local_startup();
2005-10-21 19:09:08 +02:00
while (TRUE)
{
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
if ((r= netdriver_receive(ANY, &m, &ipc_status)) != OK)
panic("dp8390: netdriver_receive failed: %d", r);
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
if (is_ipc_notify(ipc_status)) {
switch (_ENDPOINT_P(m.m_source)) {
case HARDWARE:
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
handle_hw_intr();
break;
case CLOCK:
printf("dp8390: notify from CLOCK\n");
break;
default:
panic("dp8390: illegal notify from: %d",
m.m_source);
}
/* done, get a new message */
continue;
}
switch (m.m_type)
{
2006-07-10 14:43:38 +02:00
case DL_WRITEV_S: do_vwrite_s(&m, FALSE); break;
case DL_READV_S: do_vread_s(&m); break;
case DL_CONF: do_init(&m); break;
case DL_GETSTAT_S: do_getstat_s(&m); break;
default:
panic("dp8390: illegal message: %d", m.m_type);
}
}
}
Basic System Event Framework (SEF) with ping and live update. SYSLIB CHANGES: - SEF must be used by every system process and is thereby part of the system library. - The framework provides a receive() interface (sef_receive) for system processes to automatically catch known system even messages and process them. - SEF provides a default behavior for each type of system event, but allows system processes to register callbacks to override the default behavior. - Custom (local to the process) or predefined (provided by SEF) callback implementations can be registered to SEF. - SEF currently includes support for 2 types of system events: 1. SEF Ping. The event occurs every time RS sends a ping to figure out whether a system process is still alive. The default callback implementation provided by SEF is to notify RS back to let it know the process is alive and kicking. 2. SEF Live update. The event occurs every time RS sends a prepare to update message to let a system process know an update is available and to prepare for it. The live update support is very basic for now. SEF only deals with verifying if the prepare state can be supported by the process, dumping the state for debugging purposes, and providing an event-driven programming model to the process to react to state changes check-in when ready to update. - SEF should be extended in the future to integrate support for more types of system events. Ideally, all the cross-cutting concerns should be integrated into SEF to avoid duplicating code and ease extensibility. Examples include: * PM notify messages primarily used at shutdown. * SYSTEM notify messages primarily used for signals. * CLOCK notify messages used for system alarms. * Debug messages. IS could still be in charge of fkey handling but would forward the debug message to the target process (e.g. PM, if the user requested debug information about PM). SEF would then catch the message and do nothing unless the process has registered an appropriate callback to deal with the event. This simplifies the programming model to print debug information, avoids duplicating code, and reduces the effort to print debug information. SYSTEM PROCESSES CHANGES: - Every system process registers SEF callbacks it needs to override the default system behavior and calls sef_startup() right after being started. - sef_startup() does almost nothing now, but will be extended in the future to support callbacks of its own to let RS control and synchronize with every system process at initialization time. - Every system process calls sef_receive() now rather than receive() directly, to let SEF handle predefined system events. RS CHANGES: - RS supports a basic single-component live update protocol now, as follows: * When an update command is issued (via "service update *"), RS notifies the target system process to prepare for a specific update state. * If the process doesn't respond back in time, the update is aborted. * When the process responds back, RS kills it and marks it for refreshing. * The process is then automatically restarted as for a buggy process and can start running again. * Live update is currently prototyped as a controlled failure.
2009-12-21 15:12:21 +01:00
/*===========================================================================*
* sef_local_startup *
*===========================================================================*/
2012-03-25 20:25:53 +02:00
static void sef_local_startup()
Basic System Event Framework (SEF) with ping and live update. SYSLIB CHANGES: - SEF must be used by every system process and is thereby part of the system library. - The framework provides a receive() interface (sef_receive) for system processes to automatically catch known system even messages and process them. - SEF provides a default behavior for each type of system event, but allows system processes to register callbacks to override the default behavior. - Custom (local to the process) or predefined (provided by SEF) callback implementations can be registered to SEF. - SEF currently includes support for 2 types of system events: 1. SEF Ping. The event occurs every time RS sends a ping to figure out whether a system process is still alive. The default callback implementation provided by SEF is to notify RS back to let it know the process is alive and kicking. 2. SEF Live update. The event occurs every time RS sends a prepare to update message to let a system process know an update is available and to prepare for it. The live update support is very basic for now. SEF only deals with verifying if the prepare state can be supported by the process, dumping the state for debugging purposes, and providing an event-driven programming model to the process to react to state changes check-in when ready to update. - SEF should be extended in the future to integrate support for more types of system events. Ideally, all the cross-cutting concerns should be integrated into SEF to avoid duplicating code and ease extensibility. Examples include: * PM notify messages primarily used at shutdown. * SYSTEM notify messages primarily used for signals. * CLOCK notify messages used for system alarms. * Debug messages. IS could still be in charge of fkey handling but would forward the debug message to the target process (e.g. PM, if the user requested debug information about PM). SEF would then catch the message and do nothing unless the process has registered an appropriate callback to deal with the event. This simplifies the programming model to print debug information, avoids duplicating code, and reduces the effort to print debug information. SYSTEM PROCESSES CHANGES: - Every system process registers SEF callbacks it needs to override the default system behavior and calls sef_startup() right after being started. - sef_startup() does almost nothing now, but will be extended in the future to support callbacks of its own to let RS control and synchronize with every system process at initialization time. - Every system process calls sef_receive() now rather than receive() directly, to let SEF handle predefined system events. RS CHANGES: - RS supports a basic single-component live update protocol now, as follows: * When an update command is issued (via "service update *"), RS notifies the target system process to prepare for a specific update state. * If the process doesn't respond back in time, the update is aborted. * When the process responds back, RS kills it and marks it for refreshing. * The process is then automatically restarted as for a buggy process and can start running again. * Live update is currently prototyped as a controlled failure.
2009-12-21 15:12:21 +01:00
{
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
/* Register init callbacks. */
sef_setcb_init_fresh(sef_cb_init_fresh);
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
sef_setcb_init_lu(sef_cb_init_fresh);
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
sef_setcb_init_restart(sef_cb_init_fresh);
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
/* Register live update callbacks. */
sef_setcb_lu_prepare(sef_cb_lu_prepare_always_ready);
sef_setcb_lu_state_isvalid(sef_cb_lu_state_isvalid_workfree);
Basic System Event Framework (SEF) with ping and live update. SYSLIB CHANGES: - SEF must be used by every system process and is thereby part of the system library. - The framework provides a receive() interface (sef_receive) for system processes to automatically catch known system even messages and process them. - SEF provides a default behavior for each type of system event, but allows system processes to register callbacks to override the default behavior. - Custom (local to the process) or predefined (provided by SEF) callback implementations can be registered to SEF. - SEF currently includes support for 2 types of system events: 1. SEF Ping. The event occurs every time RS sends a ping to figure out whether a system process is still alive. The default callback implementation provided by SEF is to notify RS back to let it know the process is alive and kicking. 2. SEF Live update. The event occurs every time RS sends a prepare to update message to let a system process know an update is available and to prepare for it. The live update support is very basic for now. SEF only deals with verifying if the prepare state can be supported by the process, dumping the state for debugging purposes, and providing an event-driven programming model to the process to react to state changes check-in when ready to update. - SEF should be extended in the future to integrate support for more types of system events. Ideally, all the cross-cutting concerns should be integrated into SEF to avoid duplicating code and ease extensibility. Examples include: * PM notify messages primarily used at shutdown. * SYSTEM notify messages primarily used for signals. * CLOCK notify messages used for system alarms. * Debug messages. IS could still be in charge of fkey handling but would forward the debug message to the target process (e.g. PM, if the user requested debug information about PM). SEF would then catch the message and do nothing unless the process has registered an appropriate callback to deal with the event. This simplifies the programming model to print debug information, avoids duplicating code, and reduces the effort to print debug information. SYSTEM PROCESSES CHANGES: - Every system process registers SEF callbacks it needs to override the default system behavior and calls sef_startup() right after being started. - sef_startup() does almost nothing now, but will be extended in the future to support callbacks of its own to let RS control and synchronize with every system process at initialization time. - Every system process calls sef_receive() now rather than receive() directly, to let SEF handle predefined system events. RS CHANGES: - RS supports a basic single-component live update protocol now, as follows: * When an update command is issued (via "service update *"), RS notifies the target system process to prepare for a specific update state. * If the process doesn't respond back in time, the update is aborted. * When the process responds back, RS kills it and marks it for refreshing. * The process is then automatically restarted as for a buggy process and can start running again. * Live update is currently prototyped as a controlled failure.
2009-12-21 15:12:21 +01:00
New RS and new signal handling for system processes. UPDATING INFO: 20100317: /usr/src/etc/system.conf updated to ignore default kernel calls: copy it (or merge it) to /etc/system.conf. The hello driver (/dev/hello) added to the distribution: # cd /usr/src/commands/scripts && make clean install # cd /dev && MAKEDEV hello KERNEL CHANGES: - Generic signal handling support. The kernel no longer assumes PM as a signal manager for every process. The signal manager of a given process can now be specified in its privilege slot. When a signal has to be delivered, the kernel performs the lookup and forwards the signal to the appropriate signal manager. PM is the default signal manager for user processes, RS is the default signal manager for system processes. To enable ptrace()ing for system processes, it is sufficient to change the default signal manager to PM. This will temporarily disable crash recovery, though. - sys_exit() is now split into sys_exit() (i.e. exit() for system processes, which generates a self-termination signal), and sys_clear() (i.e. used by PM to ask the kernel to clear a process slot when a process exits). - Added a new kernel call (i.e. sys_update()) to swap two process slots and implement live update. PM CHANGES: - Posix signal handling is no longer allowed for system processes. System signals are split into two fixed categories: termination and non-termination signals. When a non-termination signaled is processed, PM transforms the signal into an IPC message and delivers the message to the system process. When a termination signal is processed, PM terminates the process. - PM no longer assumes itself as the signal manager for system processes. It now makes sure that every system signal goes through the kernel before being actually processes. The kernel will then dispatch the signal to the appropriate signal manager which may or may not be PM. SYSLIB CHANGES: - Simplified SEF init and LU callbacks. - Added additional predefined SEF callbacks to debug crash recovery and live update. - Fixed a temporary ack in the SEF init protocol. SEF init reply is now completely synchronous. - Added SEF signal event type to provide a uniform interface for system processes to deal with signals. A sef_cb_signal_handler() callback is available for system processes to handle every received signal. A sef_cb_signal_manager() callback is used by signal managers to process system signals on behalf of the kernel. - Fixed a few bugs with memory mapping and DS. VM CHANGES: - Page faults and memory requests coming from the kernel are now implemented using signals. - Added a new VM call to swap two process slots and implement live update. - The call is used by RS at update time and in turn invokes the kernel call sys_update(). RS CHANGES: - RS has been reworked with a better functional decomposition. - Better kernel call masks. com.h now defines the set of very basic kernel calls every system service is allowed to use. This makes system.conf simpler and easier to maintain. In addition, this guarantees a higher level of isolation for system libraries that use one or more kernel calls internally (e.g. printf). - RS is the default signal manager for system processes. By default, RS intercepts every signal delivered to every system process. This makes crash recovery possible before bringing PM and friends in the loop. - RS now supports fast rollback when something goes wrong while initializing the new version during a live update. - Live update is now implemented by keeping the two versions side-by-side and swapping the process slots when the old version is ready to update. - Crash recovery is now implemented by keeping the two versions side-by-side and cleaning up the old version only when the recovery process is complete. DS CHANGES: - Fixed a bug when the process doing ds_publish() or ds_delete() is not known by DS. - Fixed the completely broken support for strings. String publishing is now implemented in the system library and simply wraps publishing of memory ranges. Ideally, we should adopt a similar approach for other data types as well. - Test suite fixed. DRIVER CHANGES: - The hello driver has been added to the Minix distribution to demonstrate basic live update and crash recovery functionalities. - Other drivers have been adapted to conform the new SEF interface.
2010-03-17 02:15:29 +01:00
/* Register signal callbacks. */
sef_setcb_signal_handler(sef_cb_signal_handler);
Basic System Event Framework (SEF) with ping and live update. SYSLIB CHANGES: - SEF must be used by every system process and is thereby part of the system library. - The framework provides a receive() interface (sef_receive) for system processes to automatically catch known system even messages and process them. - SEF provides a default behavior for each type of system event, but allows system processes to register callbacks to override the default behavior. - Custom (local to the process) or predefined (provided by SEF) callback implementations can be registered to SEF. - SEF currently includes support for 2 types of system events: 1. SEF Ping. The event occurs every time RS sends a ping to figure out whether a system process is still alive. The default callback implementation provided by SEF is to notify RS back to let it know the process is alive and kicking. 2. SEF Live update. The event occurs every time RS sends a prepare to update message to let a system process know an update is available and to prepare for it. The live update support is very basic for now. SEF only deals with verifying if the prepare state can be supported by the process, dumping the state for debugging purposes, and providing an event-driven programming model to the process to react to state changes check-in when ready to update. - SEF should be extended in the future to integrate support for more types of system events. Ideally, all the cross-cutting concerns should be integrated into SEF to avoid duplicating code and ease extensibility. Examples include: * PM notify messages primarily used at shutdown. * SYSTEM notify messages primarily used for signals. * CLOCK notify messages used for system alarms. * Debug messages. IS could still be in charge of fkey handling but would forward the debug message to the target process (e.g. PM, if the user requested debug information about PM). SEF would then catch the message and do nothing unless the process has registered an appropriate callback to deal with the event. This simplifies the programming model to print debug information, avoids duplicating code, and reduces the effort to print debug information. SYSTEM PROCESSES CHANGES: - Every system process registers SEF callbacks it needs to override the default system behavior and calls sef_startup() right after being started. - sef_startup() does almost nothing now, but will be extended in the future to support callbacks of its own to let RS control and synchronize with every system process at initialization time. - Every system process calls sef_receive() now rather than receive() directly, to let SEF handle predefined system events. RS CHANGES: - RS supports a basic single-component live update protocol now, as follows: * When an update command is issued (via "service update *"), RS notifies the target system process to prepare for a specific update state. * If the process doesn't respond back in time, the update is aborted. * When the process responds back, RS kills it and marks it for refreshing. * The process is then automatically restarted as for a buggy process and can start running again. * Live update is currently prototyped as a controlled failure.
2009-12-21 15:12:21 +01:00
/* Let SEF perform startup. */
sef_startup();
}
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
/*===========================================================================*
* sef_cb_init_fresh *
*===========================================================================*/
2012-03-25 20:25:53 +02:00
static int sef_cb_init_fresh(int type, sef_init_info_t *UNUSED(info))
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
{
/* Initialize the dp8390 driver. */
dpeth_t *dep;
long v;
system_hz = sys_hz();
if (env_argc < 1) {
panic("A head which at this time has no name");
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
}
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
v = 0;
(void) env_parse("instance", "d", 0, &v, 0, 255);
de_instance = (int) v;
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep = &de_state;
2012-08-02 18:54:47 +02:00
strlcpy(dep->de_name, "dp8390#0", sizeof(dep->de_name));
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep->de_name[7] += de_instance;
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
Driver refactory for live update and crash recovery. SYSLIB CHANGES: - DS calls to publish / retrieve labels consider endpoints instead of u32_t. VFS CHANGES: - mapdriver() only adds an entry in the dmap table in VFS. - dev_up() is only executed upon reception of a driver up event. INET CHANGES: - INET no longer searches for existing drivers instances at startup. - A newtwork driver is (re)initialized upon reception of a driver up event. - Networking startup is now race-free by design. No need to waste 5 seconds at startup any more. DRIVER CHANGES: - Every driver publishes driver up events when starting for the first time or in case of restart when recovery actions must be taken in the upper layers. - Driver up events are published by drivers through DS. - For regular drivers, VFS is normally the only subscriber, but not necessarily. For instance, when the filter driver is in use, it must subscribe to driver up events to initiate recovery. - For network drivers, inet is the only subscriber for now. - Every VFS driver is statically linked with libdriver, every network driver is statically linked with libnetdriver. DRIVER LIBRARIES CHANGES: - Libdriver is extended to provide generic receive() and ds_publish() interfaces for VFS drivers. - driver_receive() is a wrapper for sef_receive() also used in driver_task() to discard spurious messages that were meant to be delivered to a previous version of the driver. - driver_receive_mq() is the same as driver_receive() but integrates support for queued messages. - driver_announce() publishes a driver up event for VFS drivers and marks the driver as initialized and expecting a DEV_OPEN message. - Libnetdriver is introduced to provide similar receive() and ds_publish() interfaces for network drivers (netdriver_announce() and netdriver_receive()). - Network drivers all support live update with no state transfer now. KERNEL CHANGES: - Added kernel call statectl for state management. Used by driver_announce() to unblock eventual callers sendrecing to the driver.
2010-04-08 15:41:35 +02:00
/* Announce we are up! */
netdriver_announce();
Initialization protocol for system services. SYSLIB CHANGES: - SEF framework now supports a new SEF Init request type from RS. 3 different callbacks are available (init_fresh, init_lu, init_restart) to specify initialization code when a service starts fresh, starts after a live update, or restarts. SYSTEM SERVICE CHANGES: - Initialization code for system services is now enclosed in a callback SEF will automatically call at init time. The return code of the callback will tell RS whether the initialization completed successfully. - Each init callback can access information passed by RS to initialize. As of now, each system service has access to the public entries of RS's system process table to gather all the information required to initialize. This design eliminates many existing or potential races at boot time and provides a uniform initialization interface to system services. The same interface will be reused for the upcoming publish/subscribe model to handle dynamic registration / deregistration of system services. VM CHANGES: - Uniform privilege management for all system services. Every service uses the same call mask format. For boot services, VM copies the call mask from init data. For dynamic services, VM still receives the call mask via rs_set_priv call that will be soon replaced by the upcoming publish/subscribe model. RS CHANGES: - The system process table has been reorganized and split into private entries and public entries. Only the latter ones are exposed to system services. - VM call masks are now entirely configured in rs/table.c - RS has now its own slot in the system process table. Only kernel tasks and user processes not included in the boot image are now left out from the system process table. - RS implements the initialization protocol for system services. - For services in the boot image, RS blocks till initialization is complete and panics when failure is reported back. Services are initialized in their order of appearance in the boot image priv table and RS blocks to implements synchronous initialization for every system service having the flag SF_SYNCH_BOOT set. - For services started dynamically, the initialization protocol is implemented as though it were the first ping for the service. In this case, if the system service fails to report back (or reports failure), RS brings the service down rather than trying to restart it.
2010-01-08 02:20:42 +01:00
return(OK);
}
New RS and new signal handling for system processes. UPDATING INFO: 20100317: /usr/src/etc/system.conf updated to ignore default kernel calls: copy it (or merge it) to /etc/system.conf. The hello driver (/dev/hello) added to the distribution: # cd /usr/src/commands/scripts && make clean install # cd /dev && MAKEDEV hello KERNEL CHANGES: - Generic signal handling support. The kernel no longer assumes PM as a signal manager for every process. The signal manager of a given process can now be specified in its privilege slot. When a signal has to be delivered, the kernel performs the lookup and forwards the signal to the appropriate signal manager. PM is the default signal manager for user processes, RS is the default signal manager for system processes. To enable ptrace()ing for system processes, it is sufficient to change the default signal manager to PM. This will temporarily disable crash recovery, though. - sys_exit() is now split into sys_exit() (i.e. exit() for system processes, which generates a self-termination signal), and sys_clear() (i.e. used by PM to ask the kernel to clear a process slot when a process exits). - Added a new kernel call (i.e. sys_update()) to swap two process slots and implement live update. PM CHANGES: - Posix signal handling is no longer allowed for system processes. System signals are split into two fixed categories: termination and non-termination signals. When a non-termination signaled is processed, PM transforms the signal into an IPC message and delivers the message to the system process. When a termination signal is processed, PM terminates the process. - PM no longer assumes itself as the signal manager for system processes. It now makes sure that every system signal goes through the kernel before being actually processes. The kernel will then dispatch the signal to the appropriate signal manager which may or may not be PM. SYSLIB CHANGES: - Simplified SEF init and LU callbacks. - Added additional predefined SEF callbacks to debug crash recovery and live update. - Fixed a temporary ack in the SEF init protocol. SEF init reply is now completely synchronous. - Added SEF signal event type to provide a uniform interface for system processes to deal with signals. A sef_cb_signal_handler() callback is available for system processes to handle every received signal. A sef_cb_signal_manager() callback is used by signal managers to process system signals on behalf of the kernel. - Fixed a few bugs with memory mapping and DS. VM CHANGES: - Page faults and memory requests coming from the kernel are now implemented using signals. - Added a new VM call to swap two process slots and implement live update. - The call is used by RS at update time and in turn invokes the kernel call sys_update(). RS CHANGES: - RS has been reworked with a better functional decomposition. - Better kernel call masks. com.h now defines the set of very basic kernel calls every system service is allowed to use. This makes system.conf simpler and easier to maintain. In addition, this guarantees a higher level of isolation for system libraries that use one or more kernel calls internally (e.g. printf). - RS is the default signal manager for system processes. By default, RS intercepts every signal delivered to every system process. This makes crash recovery possible before bringing PM and friends in the loop. - RS now supports fast rollback when something goes wrong while initializing the new version during a live update. - Live update is now implemented by keeping the two versions side-by-side and swapping the process slots when the old version is ready to update. - Crash recovery is now implemented by keeping the two versions side-by-side and cleaning up the old version only when the recovery process is complete. DS CHANGES: - Fixed a bug when the process doing ds_publish() or ds_delete() is not known by DS. - Fixed the completely broken support for strings. String publishing is now implemented in the system library and simply wraps publishing of memory ranges. Ideally, we should adopt a similar approach for other data types as well. - Test suite fixed. DRIVER CHANGES: - The hello driver has been added to the Minix distribution to demonstrate basic live update and crash recovery functionalities. - Other drivers have been adapted to conform the new SEF interface.
2010-03-17 02:15:29 +01:00
/*===========================================================================*
* sef_cb_signal_handler *
*===========================================================================*/
2012-03-25 20:25:53 +02:00
static void sef_cb_signal_handler(int signo)
New RS and new signal handling for system processes. UPDATING INFO: 20100317: /usr/src/etc/system.conf updated to ignore default kernel calls: copy it (or merge it) to /etc/system.conf. The hello driver (/dev/hello) added to the distribution: # cd /usr/src/commands/scripts && make clean install # cd /dev && MAKEDEV hello KERNEL CHANGES: - Generic signal handling support. The kernel no longer assumes PM as a signal manager for every process. The signal manager of a given process can now be specified in its privilege slot. When a signal has to be delivered, the kernel performs the lookup and forwards the signal to the appropriate signal manager. PM is the default signal manager for user processes, RS is the default signal manager for system processes. To enable ptrace()ing for system processes, it is sufficient to change the default signal manager to PM. This will temporarily disable crash recovery, though. - sys_exit() is now split into sys_exit() (i.e. exit() for system processes, which generates a self-termination signal), and sys_clear() (i.e. used by PM to ask the kernel to clear a process slot when a process exits). - Added a new kernel call (i.e. sys_update()) to swap two process slots and implement live update. PM CHANGES: - Posix signal handling is no longer allowed for system processes. System signals are split into two fixed categories: termination and non-termination signals. When a non-termination signaled is processed, PM transforms the signal into an IPC message and delivers the message to the system process. When a termination signal is processed, PM terminates the process. - PM no longer assumes itself as the signal manager for system processes. It now makes sure that every system signal goes through the kernel before being actually processes. The kernel will then dispatch the signal to the appropriate signal manager which may or may not be PM. SYSLIB CHANGES: - Simplified SEF init and LU callbacks. - Added additional predefined SEF callbacks to debug crash recovery and live update. - Fixed a temporary ack in the SEF init protocol. SEF init reply is now completely synchronous. - Added SEF signal event type to provide a uniform interface for system processes to deal with signals. A sef_cb_signal_handler() callback is available for system processes to handle every received signal. A sef_cb_signal_manager() callback is used by signal managers to process system signals on behalf of the kernel. - Fixed a few bugs with memory mapping and DS. VM CHANGES: - Page faults and memory requests coming from the kernel are now implemented using signals. - Added a new VM call to swap two process slots and implement live update. - The call is used by RS at update time and in turn invokes the kernel call sys_update(). RS CHANGES: - RS has been reworked with a better functional decomposition. - Better kernel call masks. com.h now defines the set of very basic kernel calls every system service is allowed to use. This makes system.conf simpler and easier to maintain. In addition, this guarantees a higher level of isolation for system libraries that use one or more kernel calls internally (e.g. printf). - RS is the default signal manager for system processes. By default, RS intercepts every signal delivered to every system process. This makes crash recovery possible before bringing PM and friends in the loop. - RS now supports fast rollback when something goes wrong while initializing the new version during a live update. - Live update is now implemented by keeping the two versions side-by-side and swapping the process slots when the old version is ready to update. - Crash recovery is now implemented by keeping the two versions side-by-side and cleaning up the old version only when the recovery process is complete. DS CHANGES: - Fixed a bug when the process doing ds_publish() or ds_delete() is not known by DS. - Fixed the completely broken support for strings. String publishing is now implemented in the system library and simply wraps publishing of memory ranges. Ideally, we should adopt a similar approach for other data types as well. - Test suite fixed. DRIVER CHANGES: - The hello driver has been added to the Minix distribution to demonstrate basic live update and crash recovery functionalities. - Other drivers have been adapted to conform the new SEF interface.
2010-03-17 02:15:29 +01:00
{
/* Only check for termination signal, ignore anything else. */
if (signo != SIGTERM) return;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
if (de_state.de_mode == DEM_ENABLED)
dp_stop(&de_state);
New RS and new signal handling for system processes. UPDATING INFO: 20100317: /usr/src/etc/system.conf updated to ignore default kernel calls: copy it (or merge it) to /etc/system.conf. The hello driver (/dev/hello) added to the distribution: # cd /usr/src/commands/scripts && make clean install # cd /dev && MAKEDEV hello KERNEL CHANGES: - Generic signal handling support. The kernel no longer assumes PM as a signal manager for every process. The signal manager of a given process can now be specified in its privilege slot. When a signal has to be delivered, the kernel performs the lookup and forwards the signal to the appropriate signal manager. PM is the default signal manager for user processes, RS is the default signal manager for system processes. To enable ptrace()ing for system processes, it is sufficient to change the default signal manager to PM. This will temporarily disable crash recovery, though. - sys_exit() is now split into sys_exit() (i.e. exit() for system processes, which generates a self-termination signal), and sys_clear() (i.e. used by PM to ask the kernel to clear a process slot when a process exits). - Added a new kernel call (i.e. sys_update()) to swap two process slots and implement live update. PM CHANGES: - Posix signal handling is no longer allowed for system processes. System signals are split into two fixed categories: termination and non-termination signals. When a non-termination signaled is processed, PM transforms the signal into an IPC message and delivers the message to the system process. When a termination signal is processed, PM terminates the process. - PM no longer assumes itself as the signal manager for system processes. It now makes sure that every system signal goes through the kernel before being actually processes. The kernel will then dispatch the signal to the appropriate signal manager which may or may not be PM. SYSLIB CHANGES: - Simplified SEF init and LU callbacks. - Added additional predefined SEF callbacks to debug crash recovery and live update. - Fixed a temporary ack in the SEF init protocol. SEF init reply is now completely synchronous. - Added SEF signal event type to provide a uniform interface for system processes to deal with signals. A sef_cb_signal_handler() callback is available for system processes to handle every received signal. A sef_cb_signal_manager() callback is used by signal managers to process system signals on behalf of the kernel. - Fixed a few bugs with memory mapping and DS. VM CHANGES: - Page faults and memory requests coming from the kernel are now implemented using signals. - Added a new VM call to swap two process slots and implement live update. - The call is used by RS at update time and in turn invokes the kernel call sys_update(). RS CHANGES: - RS has been reworked with a better functional decomposition. - Better kernel call masks. com.h now defines the set of very basic kernel calls every system service is allowed to use. This makes system.conf simpler and easier to maintain. In addition, this guarantees a higher level of isolation for system libraries that use one or more kernel calls internally (e.g. printf). - RS is the default signal manager for system processes. By default, RS intercepts every signal delivered to every system process. This makes crash recovery possible before bringing PM and friends in the loop. - RS now supports fast rollback when something goes wrong while initializing the new version during a live update. - Live update is now implemented by keeping the two versions side-by-side and swapping the process slots when the old version is ready to update. - Crash recovery is now implemented by keeping the two versions side-by-side and cleaning up the old version only when the recovery process is complete. DS CHANGES: - Fixed a bug when the process doing ds_publish() or ds_delete() is not known by DS. - Fixed the completely broken support for strings. String publishing is now implemented in the system library and simply wraps publishing of memory ranges. Ideally, we should adopt a similar approach for other data types as well. - Test suite fixed. DRIVER CHANGES: - The hello driver has been added to the Minix distribution to demonstrate basic live update and crash recovery functionalities. - Other drivers have been adapted to conform the new SEF interface.
2010-03-17 02:15:29 +01:00
}
#if 0
/*===========================================================================*
* dp8390_dump *
*===========================================================================*/
void dp8390_dump()
{
dpeth_t *dep;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
int isr;
dep = &de_state;
printf("\n");
#if XXX
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
if (dep->de_mode == DEM_DISABLED)
printf("dp8390 instance %d is disabled\n", de_instance);
else if (dep->de_mode == DEM_SINK)
printf("dp8390 instance %d is in sink mode\n", de_instance);
#endif
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
if (dep->de_mode != DEM_ENABLED)
return;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
printf("dp8390 statistics of instance %d:\n", de_instance);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
printf("recvErr :%8ld\t", dep->de_stat.ets_recvErr);
printf("sendErr :%8ld\t", dep->de_stat.ets_sendErr);
printf("OVW :%8ld\n", dep->de_stat.ets_OVW);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
printf("CRCerr :%8ld\t", dep->de_stat.ets_CRCerr);
printf("frameAll :%8ld\t", dep->de_stat.ets_frameAll);
printf("missedP :%8ld\n", dep->de_stat.ets_missedP);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
printf("packetR :%8ld\t", dep->de_stat.ets_packetR);
printf("packetT :%8ld\t", dep->de_stat.ets_packetT);
printf("transDef :%8ld\n", dep->de_stat.ets_transDef);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
printf("collision :%8ld\t", dep->de_stat.ets_collision);
printf("transAb :%8ld\t", dep->de_stat.ets_transAb);
printf("carrSense :%8ld\n", dep->de_stat.ets_carrSense);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
printf("fifoUnder :%8ld\t", dep->de_stat.ets_fifoUnder);
printf("fifoOver :%8ld\t", dep->de_stat.ets_fifoOver);
printf("CDheartbeat:%8ld\n", dep->de_stat.ets_CDheartbeat);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
printf("OWC :%8ld\t", dep->de_stat.ets_OWC);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
isr= inb_reg0(dep, DP_ISR);
printf("dp_isr = 0x%x + 0x%x, de_flags = 0x%x\n", isr,
inb_reg0(dep, DP_ISR), dep->de_flags);
}
#endif
#if ENABLE_PCI
/*===========================================================================*
* pci_conf *
*===========================================================================*/
static void pci_conf()
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
char envvar[16];
struct dpeth *dep;
int i, pci_instance;
static int first_time= 1;
if (!first_time)
return;
first_time= 0;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep= &de_state;
2012-08-02 18:54:47 +02:00
strlcpy(envvar, "DPETH0", sizeof(envvar));
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
envvar[5] += de_instance;
if (!(dep->de_pci= env_prefix(envvar, "pci")))
return; /* no PCI config */
/* Count the number of dp instances before this one that are configured
* for PCI, so that we can skip that many when enumerating PCI devices.
*/
pci_instance= 0;
for (i= 0; i < de_instance; i++) {
envvar[5]= i;
if (env_prefix(envvar, "pci"))
pci_instance++;
}
if (!rtl_probe(dep, pci_instance))
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep->de_pci= -1;
}
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
#endif /* ENABLE_PCI */
2006-07-10 14:43:38 +02:00
/*===========================================================================*
* do_vwrite_s *
*===========================================================================*/
static void do_vwrite_s(mp, from_int)
message *mp;
int from_int;
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
int count, size;
2006-07-10 14:43:38 +02:00
int sendq_head;
dpeth_t *dep;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep= &de_state;
2006-07-10 14:43:38 +02:00
count = mp->DL_COUNT;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep->de_client= mp->m_source;
2006-07-10 14:43:38 +02:00
if (dep->de_mode == DEM_SINK)
{
assert(!from_int);
dep->de_flags |= DEF_PACK_SEND;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply(dep);
2006-07-10 14:43:38 +02:00
return;
}
assert(dep->de_mode == DEM_ENABLED);
assert(dep->de_flags & DEF_ENABLED);
if (dep->de_flags & DEF_SEND_AVAIL)
panic("dp8390: send already in progress");
2006-07-10 14:43:38 +02:00
sendq_head= dep->de_sendq_head;
if (dep->de_sendq[sendq_head].sq_filled)
{
if (from_int)
panic("dp8390: should not be sending");
2006-07-10 14:43:38 +02:00
dep->de_sendmsg= *mp;
dep->de_flags |= DEF_SEND_AVAIL;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply(dep);
2006-07-10 14:43:38 +02:00
return;
}
assert(!(dep->de_flags & DEF_PACK_SEND));
get_userdata_s(mp->m_source, mp->DL_GRANT, 0,
2006-07-10 14:43:38 +02:00
(count > IOVEC_NR ? IOVEC_NR : count) *
sizeof(dep->de_write_iovec_s.iod_iovec[0]),
dep->de_write_iovec_s.iod_iovec);
dep->de_write_iovec_s.iod_iovec_s = count;
dep->de_write_iovec_s.iod_proc_nr = mp->m_source;
2006-07-10 14:43:38 +02:00
dep->de_write_iovec_s.iod_grant = mp->DL_GRANT;
dep->de_write_iovec_s.iod_iovec_offset = 0;
dep->de_tmp_iovec_s = dep->de_write_iovec_s;
size = calc_iovec_size_s(&dep->de_tmp_iovec_s);
if (size < ETH_MIN_PACK_SIZE || size > ETH_MAX_PACK_SIZE_TAGGED)
{
panic("dp8390: invalid packet size: %d", size);
2006-07-10 14:43:38 +02:00
}
(dep->de_user2nicf_s)(dep, &dep->de_write_iovec_s, 0,
dep->de_sendq[sendq_head].sq_sendpage * DP_PAGESIZE,
size);
dep->de_sendq[sendq_head].sq_filled= TRUE;
if (dep->de_sendq_tail == sendq_head)
{
outb_reg0(dep, DP_TPSR, dep->de_sendq[sendq_head].sq_sendpage);
outb_reg0(dep, DP_TBCR1, size >> 8);
outb_reg0(dep, DP_TBCR0, size & 0xff);
outb_reg0(dep, DP_CR, CR_TXP | CR_EXTRA);/* there it goes.. */
}
else
dep->de_sendq[sendq_head].sq_size= size;
if (++sendq_head == dep->de_sendq_nr)
sendq_head= 0;
assert(sendq_head < SENDQ_NR);
dep->de_sendq_head= sendq_head;
dep->de_flags |= DEF_PACK_SEND;
/* If the interrupt handler called, don't send a reply. The reply
* will be sent after all interrupts are handled.
*/
if (from_int)
return;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply(dep);
assert(dep->de_mode == DEM_ENABLED);
assert(dep->de_flags & DEF_ENABLED);
}
2006-07-10 14:43:38 +02:00
/*===========================================================================*
* do_vread_s *
*===========================================================================*/
static void do_vread_s(mp)
message *mp;
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
int count;
2006-07-10 14:43:38 +02:00
int size;
dpeth_t *dep;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep= &de_state;
2006-07-10 14:43:38 +02:00
count = mp->DL_COUNT;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep->de_client= mp->m_source;
2006-07-10 14:43:38 +02:00
if (dep->de_mode == DEM_SINK)
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply(dep);
2006-07-10 14:43:38 +02:00
return;
}
assert(dep->de_mode == DEM_ENABLED);
assert(dep->de_flags & DEF_ENABLED);
if(dep->de_flags & DEF_READING)
panic("dp8390: read already in progress");
2006-07-10 14:43:38 +02:00
get_userdata_s(mp->m_source, mp->DL_GRANT, 0,
2006-07-10 14:43:38 +02:00
(count > IOVEC_NR ? IOVEC_NR : count) *
sizeof(dep->de_read_iovec_s.iod_iovec[0]),
dep->de_read_iovec_s.iod_iovec);
dep->de_read_iovec_s.iod_iovec_s = count;
dep->de_read_iovec_s.iod_proc_nr = mp->m_source;
2006-07-10 14:43:38 +02:00
dep->de_read_iovec_s.iod_grant = mp->DL_GRANT;
dep->de_read_iovec_s.iod_iovec_offset = 0;
dep->de_tmp_iovec_s = dep->de_read_iovec_s;
size= calc_iovec_size_s(&dep->de_tmp_iovec_s);
if (size < ETH_MAX_PACK_SIZE_TAGGED)
panic("dp8390: wrong packet size: %d", size);
2006-07-10 14:43:38 +02:00
dep->de_flags |= DEF_READING;
dp_recv(dep);
if ((dep->de_flags & (DEF_READING|DEF_STOPPED)) ==
(DEF_READING|DEF_STOPPED))
{
/* The chip is stopped, and all arrived packets are
* delivered.
*/
dp_reset(dep);
}
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply(dep);
2006-07-10 14:43:38 +02:00
}
/*===========================================================================*
* do_init *
*===========================================================================*/
static void do_init(message *mp)
{
dpeth_t *dep;
message reply_mess;
#if ENABLE_PCI
pci_conf(); /* Configure PCI devices. */
#endif
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep= &de_state;
if (dep->de_mode == DEM_DISABLED)
{
/* This is the default, try to (re)locate the device. */
conf_hw(dep);
if (dep->de_mode == DEM_DISABLED)
{
/* Probe failed, or the device is configured off. */
2006-07-10 14:43:38 +02:00
reply_mess.m_type= DL_CONF_REPLY;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply_mess.DL_STAT= ENXIO;
mess_reply(mp, &reply_mess);
return;
}
if (dep->de_mode == DEM_ENABLED)
dp_init(dep);
}
if (dep->de_mode == DEM_SINK)
{
strncpy((char *) dep->de_address.ea_addr, "ZDP", 6);
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep->de_address.ea_addr[5] = de_instance;
dp_confaddr(dep);
2006-07-10 14:43:38 +02:00
reply_mess.m_type = DL_CONF_REPLY;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply_mess.DL_STAT = OK;
*(ether_addr_t *) reply_mess.DL_HWADDR = dep->de_address;
mess_reply(mp, &reply_mess);
return;
}
assert(dep->de_mode == DEM_ENABLED);
assert(dep->de_flags & DEF_ENABLED);
dep->de_flags &= ~(DEF_PROMISC | DEF_MULTI | DEF_BROAD);
if (mp->DL_MODE & DL_PROMISC_REQ)
dep->de_flags |= DEF_PROMISC | DEF_MULTI | DEF_BROAD;
if (mp->DL_MODE & DL_MULTI_REQ)
dep->de_flags |= DEF_MULTI;
if (mp->DL_MODE & DL_BROAD_REQ)
dep->de_flags |= DEF_BROAD;
dp_reinit(dep);
2006-07-10 14:43:38 +02:00
reply_mess.m_type = DL_CONF_REPLY;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply_mess.DL_STAT = OK;
*(ether_addr_t *) reply_mess.DL_HWADDR = dep->de_address;
mess_reply(mp, &reply_mess);
}
/*===========================================================================*
* do_int *
*===========================================================================*/
static void do_int(dep)
dpeth_t *dep;
{
if (dep->de_flags & (DEF_PACK_SEND | DEF_PACK_RECV))
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply(dep);
}
2006-07-10 14:43:38 +02:00
/*===========================================================================*
* do_getstat_s *
*===========================================================================*/
static void do_getstat_s(mp)
message *mp;
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
int r;
2006-07-10 14:43:38 +02:00
dpeth_t *dep;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
dep= &de_state;
2006-07-10 14:43:38 +02:00
if (dep->de_mode == DEM_SINK)
{
put_userdata_s(mp->m_source, (vir_bytes) mp->DL_GRANT,
2006-07-10 14:43:38 +02:00
(vir_bytes) sizeof(dep->de_stat), &dep->de_stat);
mp->m_type= DL_STAT_REPLY;
r= send(mp->m_source, mp);
if (r != OK)
panic("do_getstat: send failed: %d", r);
2006-07-10 14:43:38 +02:00
return;
}
assert(dep->de_mode == DEM_ENABLED);
assert(dep->de_flags & DEF_ENABLED);
dep->de_stat.ets_CRCerr += inb_reg0(dep, DP_CNTR0);
dep->de_stat.ets_frameAll += inb_reg0(dep, DP_CNTR1);
dep->de_stat.ets_missedP += inb_reg0(dep, DP_CNTR2);
put_userdata_s(mp->m_source, mp->DL_GRANT,
2006-07-10 14:43:38 +02:00
sizeof(dep->de_stat), &dep->de_stat);
mp->m_type= DL_STAT_REPLY;
r= send(mp->m_source, mp);
if (r != OK)
panic("do_getstat: send failed: %d", r);
2006-07-10 14:43:38 +02:00
}
2005-10-21 19:09:08 +02:00
/*===========================================================================*
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
* dp_stop *
*===========================================================================*/
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
static void dp_stop(dep)
dpeth_t *dep;
{
if (dep->de_mode == DEM_SINK)
return;
assert(dep->de_mode == DEM_ENABLED);
if (!(dep->de_flags & DEF_ENABLED))
return;
outb_reg0(dep, DP_CR, CR_STP | CR_DM_ABORT);
(dep->de_stopf)(dep);
dep->de_flags= DEF_EMPTY;
}
/*===========================================================================*
* dp_init *
*===========================================================================*/
static void dp_init(dep)
dpeth_t *dep;
{
int dp_rcr_reg;
int i, r;
/* General initialization */
dep->de_flags = DEF_EMPTY;
(*dep->de_initf)(dep);
dp_confaddr(dep);
if (debug)
{
printf("%s: Ethernet address ", dep->de_name);
for (i= 0; i < 6; i++)
printf("%x%c", dep->de_address.ea_addr[i],
i < 5 ? ':' : '\n');
}
2006-07-10 14:43:38 +02:00
/* Map buffer */
map_hw_buffer(dep);
/* Initialization of the dp8390 following the mandatory procedure
* in reference manual ("DP8390D/NS32490D NIC Network Interface
* Controller", National Semiconductor, July 1995, Page 29).
*/
/* Step 1: */
outb_reg0(dep, DP_CR, CR_PS_P0 | CR_STP | CR_DM_ABORT);
/* Step 2: */
if (dep->de_16bit)
outb_reg0(dep, DP_DCR, DCR_WORDWIDE | DCR_8BYTES | DCR_BMS);
else
outb_reg0(dep, DP_DCR, DCR_BYTEWIDE | DCR_8BYTES | DCR_BMS);
/* Step 3: */
outb_reg0(dep, DP_RBCR0, 0);
outb_reg0(dep, DP_RBCR1, 0);
/* Step 4: */
dp_rcr_reg = 0;
if (dep->de_flags & DEF_PROMISC)
dp_rcr_reg |= RCR_AB | RCR_PRO | RCR_AM;
if (dep->de_flags & DEF_BROAD)
dp_rcr_reg |= RCR_AB;
if (dep->de_flags & DEF_MULTI)
dp_rcr_reg |= RCR_AM;
outb_reg0(dep, DP_RCR, dp_rcr_reg);
/* Step 5: */
outb_reg0(dep, DP_TCR, TCR_INTERNAL);
/* Step 6: */
outb_reg0(dep, DP_BNRY, dep->de_startpage);
outb_reg0(dep, DP_PSTART, dep->de_startpage);
outb_reg0(dep, DP_PSTOP, dep->de_stoppage);
/* Step 7: */
outb_reg0(dep, DP_ISR, 0xFF);
/* Step 8: */
outb_reg0(dep, DP_IMR, IMR_PRXE | IMR_PTXE | IMR_RXEE | IMR_TXEE |
IMR_OVWE | IMR_CNTE);
/* Step 9: */
outb_reg0(dep, DP_CR, CR_PS_P1 | CR_DM_ABORT | CR_STP);
outb_reg1(dep, DP_PAR0, dep->de_address.ea_addr[0]);
outb_reg1(dep, DP_PAR1, dep->de_address.ea_addr[1]);
outb_reg1(dep, DP_PAR2, dep->de_address.ea_addr[2]);
outb_reg1(dep, DP_PAR3, dep->de_address.ea_addr[3]);
outb_reg1(dep, DP_PAR4, dep->de_address.ea_addr[4]);
outb_reg1(dep, DP_PAR5, dep->de_address.ea_addr[5]);
outb_reg1(dep, DP_MAR0, 0xff);
outb_reg1(dep, DP_MAR1, 0xff);
outb_reg1(dep, DP_MAR2, 0xff);
outb_reg1(dep, DP_MAR3, 0xff);
outb_reg1(dep, DP_MAR4, 0xff);
outb_reg1(dep, DP_MAR5, 0xff);
outb_reg1(dep, DP_MAR6, 0xff);
outb_reg1(dep, DP_MAR7, 0xff);
outb_reg1(dep, DP_CURR, dep->de_startpage + 1);
/* Step 10: */
outb_reg0(dep, DP_CR, CR_DM_ABORT | CR_STA);
/* Step 11: */
outb_reg0(dep, DP_TCR, TCR_NORMAL);
inb_reg0(dep, DP_CNTR0); /* reset counters by reading */
inb_reg0(dep, DP_CNTR1);
inb_reg0(dep, DP_CNTR2);
/* Finish the initialization. */
dep->de_flags |= DEF_ENABLED;
for (i= 0; i<dep->de_sendq_nr; i++)
dep->de_sendq[i].sq_filled= 0;
dep->de_sendq_head= 0;
dep->de_sendq_tail= 0;
if (!dep->de_prog_IO)
{
2006-07-10 14:43:38 +02:00
dep->de_user2nicf_s= dp_user2nic_s;
dep->de_nic2userf_s= dp_nic2user_s;
dep->de_getblockf= dp_getblock;
}
else if (dep->de_16bit)
{
2006-07-10 14:43:38 +02:00
dep->de_user2nicf_s= dp_pio16_user2nic_s;
dep->de_nic2userf_s= dp_pio16_nic2user_s;
dep->de_getblockf= dp_pio16_getblock;
}
else
{
2006-07-10 14:43:38 +02:00
dep->de_user2nicf_s= dp_pio8_user2nic_s;
dep->de_nic2userf_s= dp_pio8_nic2user_s;
dep->de_getblockf= dp_pio8_getblock;
}
/* Set the interrupt handler and policy. Do not automatically
* reenable interrupts. Return the IRQ line number on interrupts.
*/
dep->de_hook = dep->de_irq;
r= sys_irqsetpolicy(dep->de_irq, 0, &dep->de_hook);
if (r != OK)
panic("sys_irqsetpolicy failed: %d", r);
r= sys_irqenable(&dep->de_hook);
if (r != OK)
{
panic("unable enable interrupts: %d", r);
}
}
/*===========================================================================*
* dp_confaddr *
*===========================================================================*/
static void dp_confaddr(dep)
dpeth_t *dep;
{
int i;
char eakey[16];
static char eafmt[]= "x:x:x:x:x:x";
long v;
/* User defined ethernet address? */
2012-08-02 18:54:47 +02:00
strlcpy(eakey, "DPETH0_EA", sizeof(eakey));
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
eakey[5] += de_instance;
for (i= 0; i < 6; i++)
{
v= dep->de_address.ea_addr[i];
if (env_parse(eakey, eafmt, i, &v, 0x00L, 0xFFL) != EP_SET)
{
break;
}
dep->de_address.ea_addr[i]= v;
}
if (i != 0 && i != 6) env_panic(eakey); /* It's all or nothing */
}
/*===========================================================================*
* dp_reinit *
*===========================================================================*/
static void dp_reinit(dep)
dpeth_t *dep;
{
int dp_rcr_reg;
outb_reg0(dep, DP_CR, CR_PS_P0 | CR_EXTRA);
dp_rcr_reg = 0;
if (dep->de_flags & DEF_PROMISC)
dp_rcr_reg |= RCR_AB | RCR_PRO | RCR_AM;
if (dep->de_flags & DEF_BROAD)
dp_rcr_reg |= RCR_AB;
if (dep->de_flags & DEF_MULTI)
dp_rcr_reg |= RCR_AM;
outb_reg0(dep, DP_RCR, dp_rcr_reg);
}
/*===========================================================================*
* dp_reset *
*===========================================================================*/
static void dp_reset(dep)
dpeth_t *dep;
{
int i;
/* Stop chip */
outb_reg0(dep, DP_CR, CR_STP | CR_DM_ABORT);
outb_reg0(dep, DP_RBCR0, 0);
outb_reg0(dep, DP_RBCR1, 0);
for (i= 0; i < 0x1000 && ((inb_reg0(dep, DP_ISR) & ISR_RST) == 0); i++)
; /* Do nothing */
outb_reg0(dep, DP_TCR, TCR_1EXTERNAL|TCR_OFST);
outb_reg0(dep, DP_CR, CR_STA|CR_DM_ABORT);
outb_reg0(dep, DP_TCR, TCR_NORMAL);
/* Acknowledge the ISR_RDC (remote dma) interrupt. */
for (i= 0; i < 0x1000 && ((inb_reg0(dep, DP_ISR) & ISR_RDC) == 0); i++)
; /* Do nothing */
outb_reg0(dep, DP_ISR, inb_reg0(dep, DP_ISR) & ~ISR_RDC);
/* Reset the transmit ring. If we were transmitting a packet, we
* pretend that the packet is processed. Higher layers will
* retransmit if the packet wasn't actually sent.
*/
dep->de_sendq_head= dep->de_sendq_tail= 0;
for (i= 0; i<dep->de_sendq_nr; i++)
dep->de_sendq[i].sq_filled= 0;
dp_send(dep);
dep->de_flags &= ~DEF_STOPPED;
}
/*===========================================================================*
* dp_check_ints *
*===========================================================================*/
static void dp_check_ints(dep)
dpeth_t *dep;
{
int isr, tsr;
int size, sendq_tail;
if (!(dep->de_flags & DEF_ENABLED))
panic("dp8390: got premature interrupt");
for(;;)
{
isr = inb_reg0(dep, DP_ISR);
if (!isr)
break;
outb_reg0(dep, DP_ISR, isr);
if (isr & (ISR_PTX|ISR_TXE))
{
if (isr & ISR_TXE)
{
#if DEBUG
{ printf("%s: got send Error\n", dep->de_name); }
#endif
dep->de_stat.ets_sendErr++;
}
else
{
tsr = inb_reg0(dep, DP_TSR);
if (tsr & TSR_PTX) dep->de_stat.ets_packetT++;
#if 0 /* Reserved in later manuals, should be ignored */
if (!(tsr & TSR_DFR))
{
/* In most (all?) implementations of
* the dp8390, this bit is set
* when the packet is not deferred
*/
dep->de_stat.ets_transDef++;
}
#endif
if (tsr & TSR_COL) dep->de_stat.ets_collision++;
if (tsr & TSR_ABT) dep->de_stat.ets_transAb++;
if (tsr & TSR_CRS) dep->de_stat.ets_carrSense++;
if (tsr & TSR_FU
&& ++dep->de_stat.ets_fifoUnder <= 10)
{
printf("%s: fifo underrun\n",
dep->de_name);
}
if (tsr & TSR_CDH
&& ++dep->de_stat.ets_CDheartbeat <= 10)
{
printf("%s: CD heart beat failure\n",
dep->de_name);
}
if (tsr & TSR_OWC) dep->de_stat.ets_OWC++;
}
sendq_tail= dep->de_sendq_tail;
if (!(dep->de_sendq[sendq_tail].sq_filled))
{
/* Software bug? */
assert(!debug);
/* Or hardware bug? */
printf(
"%s: transmit interrupt, but not sending\n",
dep->de_name);
continue;
}
dep->de_sendq[sendq_tail].sq_filled= 0;
if (++sendq_tail == dep->de_sendq_nr)
sendq_tail= 0;
dep->de_sendq_tail= sendq_tail;
if (dep->de_sendq[sendq_tail].sq_filled)
{
size= dep->de_sendq[sendq_tail].sq_size;
outb_reg0(dep, DP_TPSR,
dep->de_sendq[sendq_tail].sq_sendpage);
outb_reg0(dep, DP_TBCR1, size >> 8);
outb_reg0(dep, DP_TBCR0, size & 0xff);
outb_reg0(dep, DP_CR, CR_TXP | CR_EXTRA);
}
if (dep->de_flags & DEF_SEND_AVAIL)
dp_send(dep);
}
if (isr & ISR_PRX)
{
/* Only call dp_recv if there is a read request */
if (dep->de_flags & DEF_READING)
dp_recv(dep);
}
if (isr & ISR_RXE) dep->de_stat.ets_recvErr++;
if (isr & ISR_CNT)
{
dep->de_stat.ets_CRCerr += inb_reg0(dep, DP_CNTR0);
dep->de_stat.ets_frameAll += inb_reg0(dep, DP_CNTR1);
dep->de_stat.ets_missedP += inb_reg0(dep, DP_CNTR2);
}
if (isr & ISR_OVW)
{
dep->de_stat.ets_OVW++;
#if 0
{ printW(); printf(
"%s: got overwrite warning\n", dep->de_name); }
#endif
if (dep->de_flags & DEF_READING)
{
printf(
"dp_check_ints: strange: overwrite warning and pending read request\n");
dp_recv(dep);
}
}
if (isr & ISR_RDC)
{
/* Nothing to do */
}
if (isr & ISR_RST)
{
/* this means we got an interrupt but the ethernet
* chip is shutdown. We set the flag DEF_STOPPED,
* and continue processing arrived packets. When the
* receive buffer is empty, we reset the dp8390.
*/
#if 0
{ printW(); printf(
"%s: NIC stopped\n", dep->de_name); }
#endif
dep->de_flags |= DEF_STOPPED;
break;
}
}
if ((dep->de_flags & (DEF_READING|DEF_STOPPED)) ==
(DEF_READING|DEF_STOPPED))
{
/* The chip is stopped, and all arrived packets are
* delivered.
*/
dp_reset(dep);
}
}
/*===========================================================================*
* dp_recv *
*===========================================================================*/
static void dp_recv(dep)
dpeth_t *dep;
{
dp_rcvhdr_t header;
unsigned pageno, curr, next;
vir_bytes length;
int packet_processed, r;
u16_t eth_type;
packet_processed = FALSE;
pageno = inb_reg0(dep, DP_BNRY) + 1;
if (pageno == dep->de_stoppage) pageno = dep->de_startpage;
do
{
outb_reg0(dep, DP_CR, CR_PS_P1 | CR_EXTRA);
curr = inb_reg1(dep, DP_CURR);
outb_reg0(dep, DP_CR, CR_PS_P0 | CR_EXTRA);
if (curr == pageno) break;
(dep->de_getblockf)(dep, pageno, (size_t)0, sizeof(header),
&header);
(dep->de_getblockf)(dep, pageno, sizeof(header) +
2*sizeof(ether_addr_t), sizeof(eth_type), &eth_type);
length = (header.dr_rbcl | (header.dr_rbch << 8)) -
sizeof(dp_rcvhdr_t);
next = header.dr_next;
if (length < ETH_MIN_PACK_SIZE ||
length > ETH_MAX_PACK_SIZE_TAGGED)
{
printf("%s: packet with strange length arrived: %d\n",
dep->de_name, (int) length);
next= curr;
}
else if (next < dep->de_startpage || next >= dep->de_stoppage)
{
printf("%s: strange next page\n", dep->de_name);
next= curr;
}
else if (header.dr_status & RSR_FO)
{
/* This is very serious, so we issue a warning and
* reset the buffers */
printf("%s: fifo overrun, resetting receive buffer\n",
dep->de_name);
dep->de_stat.ets_fifoOver++;
next = curr;
}
else if ((header.dr_status & RSR_PRX) &&
(dep->de_flags & DEF_ENABLED))
{
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
r = dp_pkt2user_s(dep, pageno, length);
if (r != OK)
return;
packet_processed = TRUE;
dep->de_stat.ets_packetR++;
}
if (next == dep->de_startpage)
outb_reg0(dep, DP_BNRY, dep->de_stoppage - 1);
else
outb_reg0(dep, DP_BNRY, next - 1);
pageno = next;
}
while (!packet_processed);
}
/*===========================================================================*
* dp_send *
*===========================================================================*/
static void dp_send(dep)
dpeth_t *dep;
{
if (!(dep->de_flags & DEF_SEND_AVAIL))
return;
dep->de_flags &= ~DEF_SEND_AVAIL;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
do_vwrite_s(&dep->de_sendmsg, TRUE);
}
/*===========================================================================*
* dp_getblock *
*===========================================================================*/
static void dp_getblock(dep, page, offset, size, dst)
dpeth_t *dep;
int page;
size_t offset;
size_t size;
void *dst;
{
offset = page * DP_PAGESIZE + offset;
2006-07-10 14:43:38 +02:00
memcpy(dst, dep->de_locmem + offset, size);
}
/*===========================================================================*
* dp_pio8_getblock *
*===========================================================================*/
static void dp_pio8_getblock(dep, page, offset, size, dst)
dpeth_t *dep;
int page;
size_t offset;
size_t size;
void *dst;
{
offset = page * DP_PAGESIZE + offset;
outb_reg0(dep, DP_RBCR0, size & 0xFF);
outb_reg0(dep, DP_RBCR1, size >> 8);
outb_reg0(dep, DP_RSAR0, offset & 0xFF);
outb_reg0(dep, DP_RSAR1, offset >> 8);
outb_reg0(dep, DP_CR, CR_DM_RR | CR_PS_P0 | CR_STA);
insb(dep->de_data_port, dst, size);
}
/*===========================================================================*
* dp_pio16_getblock *
*===========================================================================*/
static void dp_pio16_getblock(dep, page, offset, size, dst)
dpeth_t *dep;
int page;
size_t offset;
size_t size;
void *dst;
{
offset = page * DP_PAGESIZE + offset;
outb_reg0(dep, DP_RBCR0, size & 0xFF);
outb_reg0(dep, DP_RBCR1, size >> 8);
outb_reg0(dep, DP_RSAR0, offset & 0xFF);
outb_reg0(dep, DP_RSAR1, offset >> 8);
outb_reg0(dep, DP_CR, CR_DM_RR | CR_PS_P0 | CR_STA);
assert (!(size & 1));
insw(dep->de_data_port, dst, size);
}
2006-07-10 14:43:38 +02:00
/*===========================================================================*
* dp_pkt2user_s *
*===========================================================================*/
static int dp_pkt2user_s(dpeth_t *dep, int page, vir_bytes length)
2006-07-10 14:43:38 +02:00
{
int last, count;
if (!(dep->de_flags & DEF_READING))
return EGENERIC;
last = page + (length - 1) / DP_PAGESIZE;
if (last >= dep->de_stoppage)
{
count = (dep->de_stoppage - page) * DP_PAGESIZE -
sizeof(dp_rcvhdr_t);
/* Save read_iovec since we need it twice. */
dep->de_tmp_iovec_s = dep->de_read_iovec_s;
(dep->de_nic2userf_s)(dep, page * DP_PAGESIZE +
sizeof(dp_rcvhdr_t), &dep->de_tmp_iovec_s, 0, count);
(dep->de_nic2userf_s)(dep, dep->de_startpage * DP_PAGESIZE,
&dep->de_read_iovec_s, count, length - count);
}
else
{
(dep->de_nic2userf_s)(dep, page * DP_PAGESIZE +
sizeof(dp_rcvhdr_t), &dep->de_read_iovec_s, 0, length);
}
dep->de_read_s = length;
dep->de_flags |= DEF_PACK_RECV;
dep->de_flags &= ~DEF_READING;
return OK;
}
/*===========================================================================*
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
* dp_user2nic_s *
*===========================================================================*/
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
static void dp_user2nic_s(dep, iovp, offset, nic_addr, count)
dpeth_t *dep;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
iovec_dat_s_t *iovp;
2006-07-10 14:43:38 +02:00
vir_bytes offset;
int nic_addr;
vir_bytes count;
{
vir_bytes vir_hw;
2006-07-10 14:43:38 +02:00
int bytes, i, r;
vir_hw = (vir_bytes)dep->de_locmem + nic_addr;
i= 0;
while (count > 0)
{
if (i >= IOVEC_NR)
{
dp_next_iovec_s(iovp);
i= 0;
continue;
}
assert(i < iovp->iod_iovec_s);
if (offset >= iovp->iod_iovec[i].iov_size)
{
offset -= iovp->iod_iovec[i].iov_size;
i++;
continue;
}
bytes = iovp->iod_iovec[i].iov_size - offset;
if (bytes > count)
bytes = count;
r= sys_safecopyfrom(iovp->iod_proc_nr,
iovp->iod_iovec[i].iov_grant, offset,
vir_hw, bytes);
if (r != OK) {
panic("dp_user2nic_s: sys_safecopyfrom failed: %d", r);
2006-07-10 14:43:38 +02:00
}
count -= bytes;
vir_hw += bytes;
offset += bytes;
}
assert(count == 0);
}
/*===========================================================================*
2006-07-10 14:43:38 +02:00
* dp_pio8_user2nic_s *
*===========================================================================*/
2006-07-10 14:43:38 +02:00
static void dp_pio8_user2nic_s(dep, iovp, offset, nic_addr, count)
dpeth_t *dep;
2006-07-10 14:43:38 +02:00
iovec_dat_s_t *iovp;
vir_bytes offset;
int nic_addr;
vir_bytes count;
{
2006-07-10 14:43:38 +02:00
int bytes, i, r;
outb_reg0(dep, DP_ISR, ISR_RDC);
2006-07-10 14:43:38 +02:00
outb_reg0(dep, DP_RBCR0, count & 0xFF);
outb_reg0(dep, DP_RBCR1, count >> 8);
outb_reg0(dep, DP_RSAR0, nic_addr & 0xFF);
outb_reg0(dep, DP_RSAR1, nic_addr >> 8);
outb_reg0(dep, DP_CR, CR_DM_RW | CR_PS_P0 | CR_STA);
i= 0;
while (count > 0)
{
if (i >= IOVEC_NR)
{
2006-07-10 14:43:38 +02:00
dp_next_iovec_s(iovp);
i= 0;
continue;
}
assert(i < iovp->iod_iovec_s);
if (offset >= iovp->iod_iovec[i].iov_size)
{
offset -= iovp->iod_iovec[i].iov_size;
i++;
continue;
}
bytes = iovp->iod_iovec[i].iov_size - offset;
if (bytes > count)
bytes = count;
2006-07-10 14:43:38 +02:00
r= sys_safe_outsb(dep->de_data_port, iovp->iod_proc_nr,
iovp->iod_iovec[i].iov_grant, offset, bytes);
if (r != OK) {
panic("dp_pio8_user2nic_s: sys_safe_outsb failed: %d",
r);
2006-07-10 14:43:38 +02:00
}
count -= bytes;
offset += bytes;
}
assert(count == 0);
for (i= 0; i<100; i++)
{
if (inb_reg0(dep, DP_ISR) & ISR_RDC)
break;
}
if (i == 100)
{
panic("dp8390: remote dma failed to complete");
2006-07-10 14:43:38 +02:00
}
}
/*===========================================================================*
* dp_pio16_user2nic_s *
*===========================================================================*/
static void dp_pio16_user2nic_s(dep, iovp, offset, nic_addr, count)
dpeth_t *dep;
iovec_dat_s_t *iovp;
vir_bytes offset;
int nic_addr;
vir_bytes count;
{
vir_bytes ecount;
cp_grant_id_t gid;
int i, r, bytes, user_proc;
u8_t two_bytes[2];
int odd_byte;
ecount= (count+1) & ~1;
odd_byte= 0;
outb_reg0(dep, DP_ISR, ISR_RDC);
outb_reg0(dep, DP_RBCR0, ecount & 0xFF);
outb_reg0(dep, DP_RBCR1, ecount >> 8);
outb_reg0(dep, DP_RSAR0, nic_addr & 0xFF);
outb_reg0(dep, DP_RSAR1, nic_addr >> 8);
outb_reg0(dep, DP_CR, CR_DM_RW | CR_PS_P0 | CR_STA);
i= 0;
while (count > 0)
{
if (i >= IOVEC_NR)
{
dp_next_iovec_s(iovp);
i= 0;
continue;
}
assert(i < iovp->iod_iovec_s);
if (offset >= iovp->iod_iovec[i].iov_size)
{
offset -= iovp->iod_iovec[i].iov_size;
i++;
continue;
}
bytes = iovp->iod_iovec[i].iov_size - offset;
if (bytes > count)
bytes = count;
user_proc= iovp->iod_proc_nr;
gid= iovp->iod_iovec[i].iov_grant;
if (odd_byte)
{
r= sys_safecopyfrom(user_proc, gid, offset,
(vir_bytes)&two_bytes[1], 1);
if (r != OK) {
panic("dp_pio16_user2nic: sys_safecopyfrom failed: %d", r);
2006-07-10 14:43:38 +02:00
}
outw(dep->de_data_port, *(u16_t *)two_bytes);
count--;
offset++;
bytes--;
odd_byte= 0;
if (!bytes)
continue;
}
ecount= bytes & ~1;
if (ecount != 0)
{
r= sys_safe_outsw(dep->de_data_port, user_proc,
gid, offset, ecount);
if (r != OK) {
panic("dp_pio16_user2nic: sys_safe_outsw failed: %d", r);
2006-07-10 14:43:38 +02:00
}
count -= ecount;
offset += ecount;
bytes -= ecount;
}
if (bytes)
{
assert(bytes == 1);
r= sys_safecopyfrom(user_proc, gid, offset,
(vir_bytes)&two_bytes[0], 1);
if (r != OK) {
panic("dp_pio16_user2nic: sys_safecopyfrom failed: %d", r);
2006-07-10 14:43:38 +02:00
}
count--;
offset++;
bytes--;
odd_byte= 1;
}
}
assert(count == 0);
if (odd_byte)
outw(dep->de_data_port, *(u16_t *)two_bytes);
for (i= 0; i<100; i++)
{
if (inb_reg0(dep, DP_ISR) & ISR_RDC)
break;
}
if (i == 100)
{
panic("dp8390: remote dma failed to complete");
2006-07-10 14:43:38 +02:00
}
}
/*===========================================================================*
* dp_nic2user_s *
*===========================================================================*/
static void dp_nic2user_s(dep, nic_addr, iovp, offset, count)
dpeth_t *dep;
int nic_addr;
iovec_dat_s_t *iovp;
vir_bytes offset;
vir_bytes count;
{
vir_bytes vir_hw;
2006-07-10 14:43:38 +02:00
int bytes, i, r;
vir_hw = (vir_bytes)dep->de_locmem + nic_addr;
i= 0;
while (count > 0)
{
if (i >= IOVEC_NR)
{
dp_next_iovec_s(iovp);
i= 0;
continue;
}
assert(i < iovp->iod_iovec_s);
if (offset >= iovp->iod_iovec[i].iov_size)
{
offset -= iovp->iod_iovec[i].iov_size;
i++;
continue;
}
bytes = iovp->iod_iovec[i].iov_size - offset;
if (bytes > count)
bytes = count;
r= sys_safecopyto(iovp->iod_proc_nr,
iovp->iod_iovec[i].iov_grant, offset,
vir_hw, bytes);
2006-07-10 14:43:38 +02:00
if (r != OK)
panic("dp_nic2user_s: sys_safecopyto failed: %d", r);
2006-07-10 14:43:38 +02:00
count -= bytes;
vir_hw += bytes;
offset += bytes;
}
assert(count == 0);
}
/*===========================================================================*
* dp_pio8_nic2user_s *
*===========================================================================*/
static void dp_pio8_nic2user_s(dep, nic_addr, iovp, offset, count)
dpeth_t *dep;
int nic_addr;
iovec_dat_s_t *iovp;
vir_bytes offset;
vir_bytes count;
{
int bytes, i, r;
outb_reg0(dep, DP_RBCR0, count & 0xFF);
outb_reg0(dep, DP_RBCR1, count >> 8);
outb_reg0(dep, DP_RSAR0, nic_addr & 0xFF);
outb_reg0(dep, DP_RSAR1, nic_addr >> 8);
outb_reg0(dep, DP_CR, CR_DM_RR | CR_PS_P0 | CR_STA);
i= 0;
while (count > 0)
{
if (i >= IOVEC_NR)
{
dp_next_iovec_s(iovp);
i= 0;
continue;
}
assert(i < iovp->iod_iovec_s);
if (offset >= iovp->iod_iovec[i].iov_size)
{
offset -= iovp->iod_iovec[i].iov_size;
i++;
continue;
}
bytes = iovp->iod_iovec[i].iov_size - offset;
if (bytes > count)
bytes = count;
r= sys_safe_insb(dep->de_data_port, iovp->iod_proc_nr,
iovp->iod_iovec[i].iov_grant, offset, bytes);
if (r != OK) {
panic("dp_pio8_nic2user_s: sys_safe_insb failed: %d", r);
2006-07-10 14:43:38 +02:00
}
count -= bytes;
offset += bytes;
}
assert(count == 0);
}
/*===========================================================================*
2006-07-10 14:43:38 +02:00
* dp_pio16_nic2user_s *
*===========================================================================*/
static void dp_pio16_nic2user_s(dep, nic_addr, iovp, offset, count)
dpeth_t *dep;
int nic_addr;
iovec_dat_s_t *iovp;
vir_bytes offset;
vir_bytes count;
{
vir_bytes ecount;
cp_grant_id_t gid;
int i, r, bytes, user_proc;
u8_t two_bytes[2];
int odd_byte;
ecount= (count+1) & ~1;
odd_byte= 0;
outb_reg0(dep, DP_RBCR0, ecount & 0xFF);
outb_reg0(dep, DP_RBCR1, ecount >> 8);
outb_reg0(dep, DP_RSAR0, nic_addr & 0xFF);
outb_reg0(dep, DP_RSAR1, nic_addr >> 8);
outb_reg0(dep, DP_CR, CR_DM_RR | CR_PS_P0 | CR_STA);
i= 0;
while (count > 0)
{
if (i >= IOVEC_NR)
{
dp_next_iovec_s(iovp);
i= 0;
continue;
}
assert(i < iovp->iod_iovec_s);
if (offset >= iovp->iod_iovec[i].iov_size)
{
offset -= iovp->iod_iovec[i].iov_size;
i++;
continue;
}
bytes = iovp->iod_iovec[i].iov_size - offset;
if (bytes > count)
bytes = count;
user_proc= iovp->iod_proc_nr;
gid= iovp->iod_iovec[i].iov_grant;
if (odd_byte)
{
r= sys_safecopyto(user_proc, gid, offset,
(vir_bytes)&two_bytes[1], 1);
if (r != OK) {
panic("dp_pio16_nic2user: sys_safecopyto failed: %d", r);
2006-07-10 14:43:38 +02:00
}
count--;
offset++;
bytes--;
odd_byte= 0;
if (!bytes)
continue;
}
ecount= bytes & ~1;
if (ecount != 0)
{
r= sys_safe_insw(dep->de_data_port, user_proc, gid,
offset, ecount);
if (r != OK) {
panic("dp_pio16_nic2user: sys_safe_insw failed: %d",
r);
2006-07-10 14:43:38 +02:00
}
count -= ecount;
offset += ecount;
bytes -= ecount;
}
if (bytes)
{
assert(bytes == 1);
*(u16_t *)two_bytes= inw(dep->de_data_port);
r= sys_safecopyto(user_proc, gid, offset,
(vir_bytes)&two_bytes[0], 1);
2006-07-10 14:43:38 +02:00
if (r != OK)
{
panic("dp_pio16_nic2user: sys_safecopyto failed: %d",
2006-07-10 14:43:38 +02:00
r);
}
count--;
offset++;
bytes--;
odd_byte= 1;
}
}
assert(count == 0);
}
/*===========================================================================*
* dp_next_iovec_s *
*===========================================================================*/
static void dp_next_iovec_s(iovp)
iovec_dat_s_t *iovp;
{
assert(iovp->iod_iovec_s > IOVEC_NR);
iovp->iod_iovec_s -= IOVEC_NR;
iovp->iod_iovec_offset += IOVEC_NR * sizeof(iovec_t);
get_userdata_s(iovp->iod_proc_nr, iovp->iod_grant,
iovp->iod_iovec_offset,
(iovp->iod_iovec_s > IOVEC_NR ? IOVEC_NR : iovp->iod_iovec_s) *
sizeof(iovp->iod_iovec[0]), iovp->iod_iovec);
}
/*===========================================================================*
* conf_hw *
*===========================================================================*/
static void conf_hw(dep)
dpeth_t *dep;
{
static eth_stat_t empty_stat = {0, 0, 0, 0, 0, 0 /* ,... */ };
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
int confnr;
dp_conf_t *dcp;
dep->de_mode= DEM_DISABLED; /* Superfluous */
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
/* Pick a default configuration for this instance. */
confnr= MIN(de_instance, DP_CONF_NR-1);
dcp= &dp_conf[confnr];
update_conf(dep, dcp);
if (dep->de_mode != DEM_ENABLED)
return;
if (!wdeth_probe(dep) && !ne_probe(dep) && !el2_probe(dep))
{
printf("%s: No ethernet card found at 0x%x\n",
dep->de_name, dep->de_base_port);
dep->de_mode= DEM_DISABLED;
return;
}
/* XXX */ if (dep->de_linmem == 0) dep->de_linmem= 0xFFFF0000;
dep->de_flags = DEF_EMPTY;
dep->de_stat = empty_stat;
}
/*===========================================================================*
* update_conf *
*===========================================================================*/
static void update_conf(dep, dcp)
dpeth_t *dep;
dp_conf_t *dcp;
{
long v;
static char dpc_fmt[] = "x:d:x:x";
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
char eckey[16];
#if ENABLE_PCI
if (dep->de_pci)
{
if (dep->de_pci == 1)
{
/* PCI device is present */
dep->de_mode= DEM_ENABLED;
}
return; /* Already configured */
}
#endif
strlcpy(eckey, "DPETH0", sizeof(eckey));
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
eckey[5] += de_instance;
/* Get the default settings and modify them from the environment. */
dep->de_mode= DEM_SINK;
v= dcp->dpc_port;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
switch (env_parse(eckey, dpc_fmt, 0, &v, 0x0000L, 0xFFFFL)) {
case EP_OFF:
dep->de_mode= DEM_DISABLED;
break;
case EP_ON:
case EP_SET:
dep->de_mode= DEM_ENABLED; /* Might become disabled if
* all probes fail */
break;
}
dep->de_base_port= v;
v= dcp->dpc_irq | DEI_DEFAULT;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
(void) env_parse(eckey, dpc_fmt, 1, &v, 0L, (long) NR_IRQ_VECTORS - 1);
dep->de_irq= v;
v= dcp->dpc_mem;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
(void) env_parse(eckey, dpc_fmt, 2, &v, 0L, 0xFFFFFL);
dep->de_linmem= v;
v= 0;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
(void) env_parse(eckey, dpc_fmt, 3, &v, 0x2000L, 0x8000L);
dep->de_ramsize= v;
}
2006-07-10 14:43:38 +02:00
/*===========================================================================*
* map_hw_buffer *
*===========================================================================*/
static void map_hw_buffer(dep)
dpeth_t *dep;
{
if (dep->de_prog_IO)
{
#if 0
printf(
"map_hw_buffer: programmed I/O, no need to map buffer\n");
#endif
dep->de_locmem = (char *)-dep->de_ramsize; /* trap errors */
return;
}
dep->de_locmem=
vm_map_phys(SELF, (void *) dep->de_linmem, dep->de_ramsize);
if (dep->de_locmem == MAP_FAILED)
panic("map_hw_buffer: vm_map_phys failed");
2006-07-10 14:43:38 +02:00
}
/*===========================================================================*
* calc_iovec_size_s *
*===========================================================================*/
static int calc_iovec_size_s(iovp)
iovec_dat_s_t *iovp;
{
/* Calculate the size of a request. Note that the iovec_dat
* structure will be unusable after calc_iovec_size_s.
*/
int size;
int i;
size= 0;
i= 0;
while (i < iovp->iod_iovec_s)
{
if (i >= IOVEC_NR)
{
dp_next_iovec_s(iovp);
i= 0;
continue;
}
size += iovp->iod_iovec[i].iov_size;
i++;
}
return size;
}
/*===========================================================================*
* reply *
*===========================================================================*/
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
static void reply(dep)
dpeth_t *dep;
{
message reply;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
int flags;
int r;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
flags = DL_NOFLAGS;
if (dep->de_flags & DEF_PACK_SEND)
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
flags |= DL_PACK_SEND;
if (dep->de_flags & DEF_PACK_RECV)
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
flags |= DL_PACK_RECV;
reply.m_type = DL_TASK_REPLY;
This patch switches the MINIX3 ethernet driver stack from a port-based model to an instance-based model. Each ethernet driver instance is now responsible for exactly one network interface card. The port field in /etc/inet.conf now acts as an instance field instead. This patch also updates the data link protocol. This update: - eliminates the concept of ports entirely; - eliminates DL_GETNAME entirely; - standardizes on using m_source for IPC and DL_ENDPT for safecopies; - removes error codes from TASK/STAT replies, as they were unused; - removes a number of other old or unused fields; - names and renames a few other fields. All ethernet drivers have been changed to: - conform to the new protocol, and exactly that; - take on an instance number based on a given "instance" argument; - skip that number of PCI devices in probe iterations; - use config tables and environment variables based on that number; - no longer be limited to a predefined maximum of cards in any way; - get rid of any leftover non-safecopy support and other ancient junk; - have a correct banner protocol figure, or none at all. Other changes: * Inet.conf is now taken to be line-based, and supports #-comments. No existing installations are expected to be affected by this. * A new, select-based asynchio library replaces the old one. Kindly contributed by Kees J. Bot. * Inet now supports use of select() on IP devices. Combined, the last two changes together speed up dhcpd considerably in the presence of multiple interfaces. * A small bug has been fixed in nonamed.
2010-05-18 00:22:53 +02:00
reply.DL_FLAGS = flags;
reply.DL_COUNT = dep->de_read_s;
r= send(dep->de_client, &reply);
if (r < 0)
panic("dp8390: send failed: %d", r);
dep->de_read_s = 0;
dep->de_flags &= ~(DEF_PACK_SEND | DEF_PACK_RECV);
}
/*===========================================================================*
* mess_reply *
*===========================================================================*/
static void mess_reply(req, reply_mess)
message *req;
message *reply_mess;
{
if (send(req->m_source, reply_mess) != OK)
panic("dp8390: unable to mess_reply");
}
2006-07-10 14:43:38 +02:00
/*===========================================================================*
* get_userdata_s *
*===========================================================================*/
static void get_userdata_s(user_proc, grant, offset, count, loc_addr)
int user_proc;
cp_grant_id_t grant;
vir_bytes offset;
vir_bytes count;
void *loc_addr;
{
int r;
r= sys_safecopyfrom(user_proc, grant, offset,
(vir_bytes)loc_addr, count);
2006-07-10 14:43:38 +02:00
if (r != OK)
panic("get_userdata: sys_safecopyfrom failed: %d", r);
2006-07-10 14:43:38 +02:00
}
/*===========================================================================*
* put_userdata_s *
*===========================================================================*/
static void put_userdata_s(user_proc, grant, count, loc_addr)
int user_proc;
cp_grant_id_t grant;
size_t count;
void *loc_addr;
{
int r;
r= sys_safecopyto(user_proc, grant, 0, (vir_bytes)loc_addr,
count);
2006-07-10 14:43:38 +02:00
if (r != OK)
panic("put_userdata: sys_safecopyto failed: %d", r);
2006-07-10 14:43:38 +02:00
}
u8_t inb(port_t port)
{
int r;
2005-10-21 19:09:08 +02:00
u32_t value;
r= sys_inb(port, &value);
if (r != OK)
{
printf("inb failed for port 0x%x\n", port);
panic("sys_inb failed: %d", r);
}
return value;
}
u16_t inw(port_t port)
{
int r;
2012-03-05 00:11:41 +01:00
u32_t value;
r= sys_inw(port, &value);
if (r != OK)
panic("sys_inw failed: %d", r);
2006-03-25 00:08:19 +01:00
return (u16_t) value;
}
void outb(port_t port, u8_t value)
{
int r;
r= sys_outb(port, value);
if (r != OK)
panic("sys_outb failed: %d", r);
}
void outw(port_t port, u16_t value)
{
int r;
r= sys_outw(port, value);
if (r != OK)
panic("sys_outw failed: %d", r);
}
static void insb(port_t port, void *buf, size_t size)
{
do_vir_insb(port, SELF, (vir_bytes)buf, size);
}
static void insw(port_t port, void *buf, size_t size)
{
do_vir_insw(port, SELF, (vir_bytes)buf, size);
}
static void do_vir_insb(port_t port, int proc, vir_bytes buf, size_t size)
{
int r;
2006-06-20 11:48:26 +02:00
r= sys_insb(port, proc, (void *) buf, size);
if (r != OK)
panic("sys_sdevio failed: %d", r);
}
static void do_vir_insw(port_t port, int proc, vir_bytes buf, size_t size)
{
int r;
2006-06-20 11:48:26 +02:00
r= sys_insw(port, proc, (void *) buf, size);
if (r != OK)
panic("sys_sdevio failed: %d", r);
}
/*
* $PchId: dp8390.c,v 1.25 2005/02/10 17:32:07 philip Exp $
*/