minix/lib/libaudiodriver/audio_fw.c

1065 lines
30 KiB
C
Raw Normal View History

/* Best viewed with tabsize 4
*
* This file contains a standard driver for audio devices.
* It supports double dma buffering and can be configured to use
* extra buffer space beside the dma buffer.
* This driver also support sub devices, which can be independently
* opened and closed.
*
* The driver supports the following operations:
*
* m_type DEVICE IO_ENDPT COUNT POSITION ADRRESS
* -----------------------------------------------------------------
* | DEV_OPEN | device | proc nr | | | |
* |-------------+---------+---------+---------+---------+---------|
* | DEV_CLOSE | device | proc nr | | | |
* |-------------+---------+---------+---------+---------+---------|
* | DEV_READ_S | device | proc nr | bytes | | buf ptr |
* |-------------+---------+---------+---------+---------+---------|
* | DEV_WRITE_S | device | proc nr | bytes | | buf ptr |
* |-------------+---------+---------+---------+---------+---------|
* | DEV_IOCTL_S | device | proc nr |func code| | buf ptr |
* |-------------+---------+---------+---------+---------+---------|
* | DEV_STATUS | | | | | |
* |-------------+---------+---------+---------+---------+---------|
* | HARD_INT | | | | | |
* |-------------+---------+---------+---------+---------+---------|
* | SIG_STOP | | | | | |
* -----------------------------------------------------------------
*
* The file contains one entry point:
*
* main: main entry when driver is brought up
*
* October 2007 Updated audio framework to work with mplayer, added
* savecopies (Pieter Hijma)
* February 2006 Updated audio framework,
* changed driver-framework relation (Peter Boonstoppel)
* November 2005 Created generic DMA driver framework (Laurens Bronwasser)
* August 24 2005 Ported audio driver to user space
* (only audio playback) (Peter Boonstoppel)
* May 20 1995 SB16 Driver: Michel R. Prevenier
*/
#include <minix/audio_fw.h>
Merge of David's ptrace branch. Summary: o Support for ptrace T_ATTACH/T_DETACH and T_SYSCALL o PM signal handling logic should now work properly, even with debuggers being present o Asynchronous PM/VFS protocol, full IPC support for senda(), and AMF_NOREPLY senda() flag DETAILS Process stop and delay call handling of PM: o Added sys_runctl() kernel call with sys_stop() and sys_resume() aliases, for PM to stop and resume a process o Added exception for sending/syscall-traced processes to sys_runctl(), and matching SIGKREADY pseudo-signal to PM o Fixed PM signal logic to deal with requests from a process after stopping it (so-called "delay calls"), using the SIGKREADY facility o Fixed various PM panics due to race conditions with delay calls versus VFS calls o Removed special PRIO_STOP priority value o Added SYS_LOCK RTS kernel flag, to stop an individual process from running while modifying its process structure Signal and debugger handling in PM: o Fixed debugger signals being dropped if a second signal arrives when the debugger has not retrieved the first one o Fixed debugger signals being sent to the debugger more than once o Fixed debugger signals unpausing process in VFS; removed PM_UNPAUSE_TR protocol message o Detached debugger signals from general signal logic and from being blocked on VFS calls, meaning that even VFS can now be traced o Fixed debugger being unable to receive more than one pending signal in one process stop o Fixed signal delivery being delayed needlessly when multiple signals are pending o Fixed wait test for tracer, which was returning for children that were not waited for o Removed second parallel pending call from PM to VFS for any process o Fixed process becoming runnable between exec() and debugger trap o Added support for notifying the debugger before the parent when a debugged child exits o Fixed debugger death causing child to remain stopped forever o Fixed consistently incorrect use of _NSIG Extensions to ptrace(): o Added T_ATTACH and T_DETACH ptrace request, to attach and detach a debugger to and from a process o Added T_SYSCALL ptrace request, to trace system calls o Added T_SETOPT ptrace request, to set trace options o Added TO_TRACEFORK trace option, to attach automatically to children of a traced process o Added TO_ALTEXEC trace option, to send SIGSTOP instead of SIGTRAP upon a successful exec() of the tracee o Extended T_GETUSER ptrace support to allow retrieving a process's priv structure o Removed T_STOP ptrace request again, as it does not help implementing debuggers properly o Added MINIX3-specific ptrace test (test42) o Added proper manual page for ptrace(2) Asynchronous PM/VFS interface: o Fixed asynchronous messages not being checked when receive() is called with an endpoint other than ANY o Added AMF_NOREPLY senda() flag, preventing such messages from satisfying the receive part of a sendrec() o Added asynsend3() that takes optional flags; asynsend() is now a #define passing in 0 as third parameter o Made PM/VFS protocol asynchronous; reintroduced tell_fs() o Made PM_BASE request/reply number range unique o Hacked in a horrible temporary workaround into RS to deal with newly revealed RS-PM-VFS race condition triangle until VFS is asynchronous System signal handling: o Fixed shutdown logic of device drivers; removed old SIGKSTOP signal o Removed is-superuser check from PM's do_procstat() (aka getsigset()) o Added sigset macros to allow system processes to deal with the full signal set, rather than just the POSIX subset Miscellaneous PM fixes: o Split do_getset into do_get and do_set, merging common code and making structure clearer o Fixed setpriority() being able to put to sleep processes using an invalid parameter, or revive zombie processes o Made find_proc() global; removed obsolete proc_from_pid() o Cleanup here and there Also included: o Fixed false-positive boot order kernel warning o Removed last traces of old NOTIFY_FROM code THINGS OF POSSIBLE INTEREST o It should now be possible to run PM at any priority, even lower than user processes o No assumptions are made about communication speed between PM and VFS, although communication must be FIFO o A debugger will now receive incoming debuggee signals at kill time only; the process may not yet be fully stopped o A first step has been made towards making the SYSTEM task preemptible
2009-09-30 11:57:22 +02:00
#include <minix/endpoint.h>
#include <minix/ds.h>
FORWARD _PROTOTYPE( int msg_open, (int minor_dev_nr) );
FORWARD _PROTOTYPE( int msg_close, (int minor_dev_nr) );
FORWARD _PROTOTYPE( int msg_ioctl, (const message *m_ptr) );
FORWARD _PROTOTYPE( void msg_write, (const message *m_ptr) );
FORWARD _PROTOTYPE( void msg_read, (message *m_ptr) );
FORWARD _PROTOTYPE( void msg_hardware, (void) );
FORWARD _PROTOTYPE( void msg_status, (message *m_ptr) );
FORWARD _PROTOTYPE( int init_driver, (void) );
FORWARD _PROTOTYPE( int open_sub_dev, (int sub_dev_nr, int operation) );
FORWARD _PROTOTYPE( int close_sub_dev, (int sub_dev_nr) );
FORWARD _PROTOTYPE( void handle_int_write,(int sub_dev_nr) );
FORWARD _PROTOTYPE( void handle_int_read,(int sub_dev_nr) );
FORWARD _PROTOTYPE( void data_to_user, (sub_dev_t *sub_dev_ptr) );
FORWARD _PROTOTYPE( void data_from_user, (sub_dev_t *sub_dev_ptr) );
FORWARD _PROTOTYPE( int init_buffers, (sub_dev_t *sub_dev_ptr) );
FORWARD _PROTOTYPE( int get_started, (sub_dev_t *sub_dev_ptr) );
FORWARD _PROTOTYPE( void reply,(int code, int replyee, int process,int status));
FORWARD _PROTOTYPE( int io_ctl_length, (int io_request) );
FORWARD _PROTOTYPE( special_file_t* get_special_file, (int minor_dev_nr) );
FORWARD _PROTOTYPE( void tell_dev, (vir_bytes buf, size_t size, int pci_bus,
int pci_dev, int pci_func) );
PRIVATE char io_ctl_buf[_IOCPARM_MASK];
PRIVATE int irq_hook_id = 0; /* id of irq hook at the kernel */
PRIVATE int irq_hook_set = FALSE;
PRIVATE int device_available = 0;/*todo*/
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 functions and variables. */
FORWARD _PROTOTYPE( void sef_local_startup, (void) );
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
FORWARD _PROTOTYPE( int sef_cb_init_fresh, (int type, sef_init_info_t *info) );
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
FORWARD _PROTOTYPE( void sef_cb_signal_handler, (int signo) );
EXTERN _PROTOTYPE( int sef_cb_lu_prepare, (int state) );
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
EXTERN _PROTOTYPE( int sef_cb_lu_state_isvalid, (int state) );
EXTERN _PROTOTYPE( void sef_cb_lu_state_dump, (int state) );
PUBLIC int is_status_msg_expected = FALSE;
PUBLIC int main(int argc, char *argv[])
{
int r, caller;
message mess, repl_mess;
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;
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. */
sef_local_startup();
/* Here is the main loop of the dma driver. It waits for a message,
carries it out, and sends a reply. */
while(1) {
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(driver_receive(ANY, &mess, &ipc_status) != OK) {
panic("driver_receive failed");
}
caller = mess.m_source;
Merge of David's ptrace branch. Summary: o Support for ptrace T_ATTACH/T_DETACH and T_SYSCALL o PM signal handling logic should now work properly, even with debuggers being present o Asynchronous PM/VFS protocol, full IPC support for senda(), and AMF_NOREPLY senda() flag DETAILS Process stop and delay call handling of PM: o Added sys_runctl() kernel call with sys_stop() and sys_resume() aliases, for PM to stop and resume a process o Added exception for sending/syscall-traced processes to sys_runctl(), and matching SIGKREADY pseudo-signal to PM o Fixed PM signal logic to deal with requests from a process after stopping it (so-called "delay calls"), using the SIGKREADY facility o Fixed various PM panics due to race conditions with delay calls versus VFS calls o Removed special PRIO_STOP priority value o Added SYS_LOCK RTS kernel flag, to stop an individual process from running while modifying its process structure Signal and debugger handling in PM: o Fixed debugger signals being dropped if a second signal arrives when the debugger has not retrieved the first one o Fixed debugger signals being sent to the debugger more than once o Fixed debugger signals unpausing process in VFS; removed PM_UNPAUSE_TR protocol message o Detached debugger signals from general signal logic and from being blocked on VFS calls, meaning that even VFS can now be traced o Fixed debugger being unable to receive more than one pending signal in one process stop o Fixed signal delivery being delayed needlessly when multiple signals are pending o Fixed wait test for tracer, which was returning for children that were not waited for o Removed second parallel pending call from PM to VFS for any process o Fixed process becoming runnable between exec() and debugger trap o Added support for notifying the debugger before the parent when a debugged child exits o Fixed debugger death causing child to remain stopped forever o Fixed consistently incorrect use of _NSIG Extensions to ptrace(): o Added T_ATTACH and T_DETACH ptrace request, to attach and detach a debugger to and from a process o Added T_SYSCALL ptrace request, to trace system calls o Added T_SETOPT ptrace request, to set trace options o Added TO_TRACEFORK trace option, to attach automatically to children of a traced process o Added TO_ALTEXEC trace option, to send SIGSTOP instead of SIGTRAP upon a successful exec() of the tracee o Extended T_GETUSER ptrace support to allow retrieving a process's priv structure o Removed T_STOP ptrace request again, as it does not help implementing debuggers properly o Added MINIX3-specific ptrace test (test42) o Added proper manual page for ptrace(2) Asynchronous PM/VFS interface: o Fixed asynchronous messages not being checked when receive() is called with an endpoint other than ANY o Added AMF_NOREPLY senda() flag, preventing such messages from satisfying the receive part of a sendrec() o Added asynsend3() that takes optional flags; asynsend() is now a #define passing in 0 as third parameter o Made PM/VFS protocol asynchronous; reintroduced tell_fs() o Made PM_BASE request/reply number range unique o Hacked in a horrible temporary workaround into RS to deal with newly revealed RS-PM-VFS race condition triangle until VFS is asynchronous System signal handling: o Fixed shutdown logic of device drivers; removed old SIGKSTOP signal o Removed is-superuser check from PM's do_procstat() (aka getsigset()) o Added sigset macros to allow system processes to deal with the full signal set, rather than just the POSIX subset Miscellaneous PM fixes: o Split do_getset into do_get and do_set, merging common code and making structure clearer o Fixed setpriority() being able to put to sleep processes using an invalid parameter, or revive zombie processes o Made find_proc() global; removed obsolete proc_from_pid() o Cleanup here and there Also included: o Fixed false-positive boot order kernel warning o Removed last traces of old NOTIFY_FROM code THINGS OF POSSIBLE INTEREST o It should now be possible to run PM at any priority, even lower than user processes o No assumptions are made about communication speed between PM and VFS, although communication must be FIFO o A debugger will now receive incoming debuggee signals at kill time only; the process may not yet be fully stopped o A first step has been made towards making the SYSTEM task preemptible
2009-09-30 11:57:22 +02:00
/* Now carry out the work. First check for notifications. */
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)) {
Merge of David's ptrace branch. Summary: o Support for ptrace T_ATTACH/T_DETACH and T_SYSCALL o PM signal handling logic should now work properly, even with debuggers being present o Asynchronous PM/VFS protocol, full IPC support for senda(), and AMF_NOREPLY senda() flag DETAILS Process stop and delay call handling of PM: o Added sys_runctl() kernel call with sys_stop() and sys_resume() aliases, for PM to stop and resume a process o Added exception for sending/syscall-traced processes to sys_runctl(), and matching SIGKREADY pseudo-signal to PM o Fixed PM signal logic to deal with requests from a process after stopping it (so-called "delay calls"), using the SIGKREADY facility o Fixed various PM panics due to race conditions with delay calls versus VFS calls o Removed special PRIO_STOP priority value o Added SYS_LOCK RTS kernel flag, to stop an individual process from running while modifying its process structure Signal and debugger handling in PM: o Fixed debugger signals being dropped if a second signal arrives when the debugger has not retrieved the first one o Fixed debugger signals being sent to the debugger more than once o Fixed debugger signals unpausing process in VFS; removed PM_UNPAUSE_TR protocol message o Detached debugger signals from general signal logic and from being blocked on VFS calls, meaning that even VFS can now be traced o Fixed debugger being unable to receive more than one pending signal in one process stop o Fixed signal delivery being delayed needlessly when multiple signals are pending o Fixed wait test for tracer, which was returning for children that were not waited for o Removed second parallel pending call from PM to VFS for any process o Fixed process becoming runnable between exec() and debugger trap o Added support for notifying the debugger before the parent when a debugged child exits o Fixed debugger death causing child to remain stopped forever o Fixed consistently incorrect use of _NSIG Extensions to ptrace(): o Added T_ATTACH and T_DETACH ptrace request, to attach and detach a debugger to and from a process o Added T_SYSCALL ptrace request, to trace system calls o Added T_SETOPT ptrace request, to set trace options o Added TO_TRACEFORK trace option, to attach automatically to children of a traced process o Added TO_ALTEXEC trace option, to send SIGSTOP instead of SIGTRAP upon a successful exec() of the tracee o Extended T_GETUSER ptrace support to allow retrieving a process's priv structure o Removed T_STOP ptrace request again, as it does not help implementing debuggers properly o Added MINIX3-specific ptrace test (test42) o Added proper manual page for ptrace(2) Asynchronous PM/VFS interface: o Fixed asynchronous messages not being checked when receive() is called with an endpoint other than ANY o Added AMF_NOREPLY senda() flag, preventing such messages from satisfying the receive part of a sendrec() o Added asynsend3() that takes optional flags; asynsend() is now a #define passing in 0 as third parameter o Made PM/VFS protocol asynchronous; reintroduced tell_fs() o Made PM_BASE request/reply number range unique o Hacked in a horrible temporary workaround into RS to deal with newly revealed RS-PM-VFS race condition triangle until VFS is asynchronous System signal handling: o Fixed shutdown logic of device drivers; removed old SIGKSTOP signal o Removed is-superuser check from PM's do_procstat() (aka getsigset()) o Added sigset macros to allow system processes to deal with the full signal set, rather than just the POSIX subset Miscellaneous PM fixes: o Split do_getset into do_get and do_set, merging common code and making structure clearer o Fixed setpriority() being able to put to sleep processes using an invalid parameter, or revive zombie processes o Made find_proc() global; removed obsolete proc_from_pid() o Cleanup here and there Also included: o Fixed false-positive boot order kernel warning o Removed last traces of old NOTIFY_FROM code THINGS OF POSSIBLE INTEREST o It should now be possible to run PM at any priority, even lower than user processes o No assumptions are made about communication speed between PM and VFS, although communication must be FIFO o A debugger will now receive incoming debuggee signals at kill time only; the process may not yet be fully stopped o A first step has been made towards making the SYSTEM task preemptible
2009-09-30 11:57:22 +02:00
switch (_ENDPOINT_P(mess.m_source)) {
case HARDWARE:
msg_hardware();
break;
default:
dprint("%s: %d uncaught notify!\n",
drv.DriverName, mess.m_type);
}
/* get next message */
continue;
}
Merge of David's ptrace branch. Summary: o Support for ptrace T_ATTACH/T_DETACH and T_SYSCALL o PM signal handling logic should now work properly, even with debuggers being present o Asynchronous PM/VFS protocol, full IPC support for senda(), and AMF_NOREPLY senda() flag DETAILS Process stop and delay call handling of PM: o Added sys_runctl() kernel call with sys_stop() and sys_resume() aliases, for PM to stop and resume a process o Added exception for sending/syscall-traced processes to sys_runctl(), and matching SIGKREADY pseudo-signal to PM o Fixed PM signal logic to deal with requests from a process after stopping it (so-called "delay calls"), using the SIGKREADY facility o Fixed various PM panics due to race conditions with delay calls versus VFS calls o Removed special PRIO_STOP priority value o Added SYS_LOCK RTS kernel flag, to stop an individual process from running while modifying its process structure Signal and debugger handling in PM: o Fixed debugger signals being dropped if a second signal arrives when the debugger has not retrieved the first one o Fixed debugger signals being sent to the debugger more than once o Fixed debugger signals unpausing process in VFS; removed PM_UNPAUSE_TR protocol message o Detached debugger signals from general signal logic and from being blocked on VFS calls, meaning that even VFS can now be traced o Fixed debugger being unable to receive more than one pending signal in one process stop o Fixed signal delivery being delayed needlessly when multiple signals are pending o Fixed wait test for tracer, which was returning for children that were not waited for o Removed second parallel pending call from PM to VFS for any process o Fixed process becoming runnable between exec() and debugger trap o Added support for notifying the debugger before the parent when a debugged child exits o Fixed debugger death causing child to remain stopped forever o Fixed consistently incorrect use of _NSIG Extensions to ptrace(): o Added T_ATTACH and T_DETACH ptrace request, to attach and detach a debugger to and from a process o Added T_SYSCALL ptrace request, to trace system calls o Added T_SETOPT ptrace request, to set trace options o Added TO_TRACEFORK trace option, to attach automatically to children of a traced process o Added TO_ALTEXEC trace option, to send SIGSTOP instead of SIGTRAP upon a successful exec() of the tracee o Extended T_GETUSER ptrace support to allow retrieving a process's priv structure o Removed T_STOP ptrace request again, as it does not help implementing debuggers properly o Added MINIX3-specific ptrace test (test42) o Added proper manual page for ptrace(2) Asynchronous PM/VFS interface: o Fixed asynchronous messages not being checked when receive() is called with an endpoint other than ANY o Added AMF_NOREPLY senda() flag, preventing such messages from satisfying the receive part of a sendrec() o Added asynsend3() that takes optional flags; asynsend() is now a #define passing in 0 as third parameter o Made PM/VFS protocol asynchronous; reintroduced tell_fs() o Made PM_BASE request/reply number range unique o Hacked in a horrible temporary workaround into RS to deal with newly revealed RS-PM-VFS race condition triangle until VFS is asynchronous System signal handling: o Fixed shutdown logic of device drivers; removed old SIGKSTOP signal o Removed is-superuser check from PM's do_procstat() (aka getsigset()) o Added sigset macros to allow system processes to deal with the full signal set, rather than just the POSIX subset Miscellaneous PM fixes: o Split do_getset into do_get and do_set, merging common code and making structure clearer o Fixed setpriority() being able to put to sleep processes using an invalid parameter, or revive zombie processes o Made find_proc() global; removed obsolete proc_from_pid() o Cleanup here and there Also included: o Fixed false-positive boot order kernel warning o Removed last traces of old NOTIFY_FROM code THINGS OF POSSIBLE INTEREST o It should now be possible to run PM at any priority, even lower than user processes o No assumptions are made about communication speed between PM and VFS, although communication must be FIFO o A debugger will now receive incoming debuggee signals at kill time only; the process may not yet be fully stopped o A first step has been made towards making the SYSTEM task preemptible
2009-09-30 11:57:22 +02:00
/* Normal messages. */
switch(mess.m_type) {
case DEV_OPEN:
/* open the special file ( = parameter) */
r = msg_open(mess.DEVICE);
repl_mess.m_type = DEV_REVIVE;
repl_mess.REP_ENDPT = mess.IO_ENDPT;
repl_mess.REP_STATUS = r;
send(caller, &repl_mess);
continue;
case DEV_CLOSE:
/* close the special file ( = parameter) */
r = msg_close(mess.DEVICE);
repl_mess.m_type = DEV_CLOSE_REPL;
repl_mess.REP_ENDPT = mess.IO_ENDPT;
repl_mess.REP_STATUS = r;
send(caller, &repl_mess);
continue;
case DEV_IOCTL_S:
r = msg_ioctl(&mess);
if (r != SUSPEND)
{
repl_mess.m_type = DEV_REVIVE;
repl_mess.REP_ENDPT = mess.IO_ENDPT;
repl_mess.REP_IO_GRANT =
(unsigned)mess.IO_GRANT;
repl_mess.REP_STATUS = r;
send(caller, &repl_mess);
}
continue;
case DEV_READ_S:
msg_read(&mess); continue; /* don't reply */
case DEV_WRITE_S:
msg_write(&mess); continue; /* don't reply */
case DEV_STATUS:
msg_status(&mess);continue; /* don't reply */
case DEV_REOPEN:
/* reopen the special file ( = parameter) */
r = msg_open(mess.DEVICE);
repl_mess.m_type = DEV_REOPEN_REPL;
repl_mess.REP_ENDPT = mess.IO_ENDPT;
repl_mess.REP_STATUS = r;
send(caller, &repl_mess);
continue;
default:
dprint("%s: %d uncaught msg!\n",
drv.DriverName, mess.m_type);
continue;
}
/* Should not be here. Just continue. */
}
return 1;
}
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 *
*===========================================================================*/
PRIVATE void sef_local_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
/* Register init callbacks. */
sef_setcb_init_fresh(sef_cb_init_fresh);
sef_setcb_init_lu(sef_cb_init_fresh);
sef_setcb_init_restart(sef_cb_init_fresh);
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
/* Register live update callbacks. */
sef_setcb_lu_prepare(sef_cb_lu_prepare);
sef_setcb_lu_state_isvalid(sef_cb_lu_state_isvalid);
sef_setcb_lu_state_dump(sef_cb_lu_state_dump);
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 *
*===========================================================================*/
PRIVATE int sef_cb_init_fresh(int UNUSED(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 audio driver framework. */
return init_driver();
}
PRIVATE int init_driver(void) {
u32_t i; char irq;
static int executed = 0;
sub_dev_t* sub_dev_ptr;
/* init variables, get dma buffers */
for (i = 0; i < drv.NrOfSubDevices; i++) {
sub_dev_ptr = &sub_dev[i];
sub_dev_ptr->Opened = FALSE;
sub_dev_ptr->DmaBusy = FALSE;
sub_dev_ptr->DmaMode = NO_DMA;
sub_dev_ptr->DmaReadNext = 0;
sub_dev_ptr->DmaFillNext = 0;
sub_dev_ptr->DmaLength = 0;
sub_dev_ptr->BufReadNext = 0;
sub_dev_ptr->BufFillNext = 0;
sub_dev_ptr->RevivePending = FALSE;
sub_dev_ptr->OutOfData = FALSE;
sub_dev_ptr->Nr = i;
}
/* initialize hardware*/
if (drv_init_hw() != OK) {
error("%s: Could not initialize hardware\n", drv.DriverName);
return EIO;
}
/* get irq from device driver...*/
if (drv_get_irq(&irq) != OK) {
error("%s: init driver couldn't get IRQ", drv.DriverName);
return EIO;
}
/* TODO: execute the rest of this function only once
we don't want to set irq policy twice */
if (executed) return OK;
executed = TRUE;
/* ...and register interrupt vector */
if ((i=sys_irqsetpolicy(irq, 0, &irq_hook_id )) != OK){
error("%s: init driver couldn't set IRQ policy: %d", drv.DriverName, i);
return EIO;
}
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
irq_hook_set = TRUE; /* now signal handler knows it must unregister policy*/
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! */
driver_announce();
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 *
*===========================================================================*/
PRIVATE void sef_cb_signal_handler(int signo)
{
int i;
char irq;
/* Only check for termination signal, ignore anything else. */
if (signo != SIGTERM) return;
for (i = 0; i < drv.NrOfSubDevices; i++) {
drv_stop(i); /* stop all sub devices */
}
if (irq_hook_set) {
if (sys_irqdisable(&irq_hook_id) != OK) {
error("Could not disable IRQ\n");
}
/* get irq from device driver*/
if (drv_get_irq(&irq) != OK) {
error("Msg SIG_STOP Couldn't get IRQ");
}
/* remove the policy */
if (sys_irqrmpolicy(&irq_hook_id) != OK) {
error("%s: Could not disable IRQ\n",drv.DriverName);
}
}
}
PRIVATE int msg_open (int minor_dev_nr) {
int r, read_chan, write_chan, io_ctl;
special_file_t* special_file_ptr;
dprint("%s: msg_open() special file %d\n", drv.DriverName, minor_dev_nr);
special_file_ptr = get_special_file(minor_dev_nr);
if(special_file_ptr == NULL) {
return EIO;
}
read_chan = special_file_ptr->read_chan;
write_chan = special_file_ptr->write_chan;
io_ctl = special_file_ptr->io_ctl;
if (read_chan==NO_CHANNEL && write_chan==NO_CHANNEL && io_ctl==NO_CHANNEL) {
error("%s: No channel specified for minor device %d!\n",
drv.DriverName, minor_dev_nr);
return EIO;
}
if (read_chan == write_chan && read_chan != NO_CHANNEL) {
error("%s: Read and write channels are equal: %d!\n",
drv.DriverName, minor_dev_nr);
return EIO;
}
/* init driver */
if (!device_available) {
if (init_driver() != OK) {
error("%s: Couldn't init driver!\n", drv.DriverName);
return EIO;
} else {
device_available = TRUE;
}
}
/* open the sub devices specified in the interface header file */
if (write_chan != NO_CHANNEL) {
/* open sub device for writing */
if (open_sub_dev(write_chan, DEV_WRITE_S) != OK) return EIO;
}
if (read_chan != NO_CHANNEL) {
if (open_sub_dev(read_chan, DEV_READ_S) != OK) return EIO;
}
if (read_chan == io_ctl || write_chan == io_ctl) {
/* io_ctl is already opened because it's the same as read or write */
return OK; /* we're done */
}
if (io_ctl != NO_CHANNEL) { /* Ioctl differs from read/write channels, */
r = open_sub_dev(io_ctl, NO_DMA); /* open it explicitly */
if (r != OK) return EIO;
}
return OK;
}
PRIVATE int open_sub_dev(int sub_dev_nr, int dma_mode) {
sub_dev_t* sub_dev_ptr;
sub_dev_ptr = &sub_dev[sub_dev_nr];
/* Only one open at a time per sub device */
if (sub_dev_ptr->Opened) {
error("%s: Sub device %d is already opened\n",
drv.DriverName, sub_dev_nr);
return EBUSY;
}
if (sub_dev_ptr->DmaBusy) {
error("%s: Sub device %d is still busy\n", drv.DriverName, sub_dev_nr);
return EBUSY;
}
/* Setup variables */
sub_dev_ptr->Opened = TRUE;
sub_dev_ptr->DmaReadNext = 0;
sub_dev_ptr->DmaFillNext = 0;
sub_dev_ptr->DmaLength = 0;
sub_dev_ptr->DmaMode = dma_mode;
sub_dev_ptr->BufReadNext = 0;
sub_dev_ptr->BufFillNext = 0;
sub_dev_ptr->BufLength = 0;
sub_dev_ptr->RevivePending = FALSE;
sub_dev_ptr->OutOfData = TRUE;
/* arrange DMA */
if (dma_mode != NO_DMA) { /* sub device uses DMA */
/* allocate dma buffer and extra buffer space
and configure sub device for dma */
if (init_buffers(sub_dev_ptr) != OK ) return EIO;
}
return OK;
}
PRIVATE int msg_close(int minor_dev_nr) {
int r, read_chan, write_chan, io_ctl;
special_file_t* special_file_ptr;
dprint("%s: msg_close() minor device %d\n", drv.DriverName, minor_dev_nr);
special_file_ptr = get_special_file(minor_dev_nr);
if(special_file_ptr == NULL) {
return EIO;
}
read_chan = special_file_ptr->read_chan;
write_chan = special_file_ptr->write_chan;
io_ctl = special_file_ptr->io_ctl;
r= OK;
/* close all sub devices */
if (write_chan != NO_CHANNEL) {
if (close_sub_dev(write_chan) != OK) r = EIO;
}
if (read_chan != NO_CHANNEL) {
if (close_sub_dev(read_chan) != OK) r = EIO;
}
if (read_chan == io_ctl || write_chan == io_ctl) {
/* io_ctl is already closed because it's the same as read or write */
return r; /* we're done */
}
/* ioctl differs from read/write channels... */
if (io_ctl != NO_CHANNEL) {
if (close_sub_dev(io_ctl) != OK) r = EIO; /* ...close it explicitly */
}
return r;
}
PRIVATE int close_sub_dev(int sub_dev_nr) {
sub_dev_t *sub_dev_ptr;
sub_dev_ptr = &sub_dev[sub_dev_nr];
if (sub_dev_ptr->DmaMode == DEV_WRITE_S && !sub_dev_ptr->OutOfData) {
/* do nothing, still data in buffers that has to be transferred */
sub_dev_ptr->Opened = FALSE; /* keep DMA busy */
return OK;
}
if (sub_dev_ptr->DmaMode == NO_DMA) {
/* do nothing, there is no dma going on */
sub_dev_ptr->Opened = FALSE;
return OK;
}
sub_dev_ptr->Opened = FALSE;
sub_dev_ptr->DmaBusy = FALSE;
/* stop the device */
drv_stop(sub_dev_ptr->Nr);
/* free the buffers */
free(sub_dev_ptr->DmaBuf);
free(sub_dev_ptr->ExtraBuf);
2010-01-11 15:22:29 +01:00
return OK;
}
PRIVATE int msg_ioctl(const message *m_ptr)
{
int status, len, chan;
sub_dev_t *sub_dev_ptr;
special_file_t* special_file_ptr;
dprint("%s: msg_ioctl() device %d\n", drv.DriverName, m_ptr->DEVICE);
special_file_ptr = get_special_file(m_ptr->DEVICE);
if(special_file_ptr == NULL) {
return EIO;
}
chan = special_file_ptr->io_ctl;
if (chan == NO_CHANNEL) {
error("%s: No io control channel specified!\n", drv.DriverName);
return EIO;
}
/* get pointer to sub device data */
sub_dev_ptr = &sub_dev[chan];
if(!sub_dev_ptr->Opened) {
error("%s: io control impossible - not opened!\n", drv.DriverName);
return EIO;
}
/* this is a hack...todo: may we intercept reset calls? */
/*
if(m_ptr->REQUEST == DSPIORESET) {
device_available = FALSE;
}
*/
if (m_ptr->REQUEST & _IOC_IN) { /* if there is data for us, copy it */
len = io_ctl_length(m_ptr->REQUEST);
if(sys_safecopyfrom(m_ptr->IO_ENDPT,
(vir_bytes)m_ptr->ADDRESS, 0,
(vir_bytes)io_ctl_buf, len, D) != OK) {
printf("%s:%d: safecopyfrom failed\n", __FILE__, __LINE__);
}
}
/* all ioctl's are passed to the device specific part of the driver */
status = drv_io_ctl(m_ptr->REQUEST, (void *)io_ctl_buf, &len, chan);
/* _IOC_OUT bit -> user expects data */
if (status == OK && m_ptr->REQUEST & _IOC_OUT) {
/* copy result back to user */
if(sys_safecopyto(m_ptr->IO_ENDPT, (vir_bytes)m_ptr->ADDRESS, 0,
(vir_bytes)io_ctl_buf, len, D) != OK) {
printf("%s:%d: safecopyto failed\n", __FILE__, __LINE__);
}
}
return status;
}
PRIVATE void msg_write(const message *m_ptr)
{
int chan; sub_dev_t *sub_dev_ptr;
special_file_t* special_file_ptr;
dprint("%s: msg_write() device %d\n", drv.DriverName, m_ptr->DEVICE);
special_file_ptr = get_special_file(m_ptr->DEVICE);
chan = special_file_ptr->write_chan;
if (chan == NO_CHANNEL) {
error("%s: No write channel specified!\n", drv.DriverName);
reply(DEV_REVIVE, m_ptr->m_source, m_ptr->IO_ENDPT, EIO);
return;
}
/* get pointer to sub device data */
sub_dev_ptr = &sub_dev[chan];
if (!sub_dev_ptr->DmaBusy) { /* get fragment size on first write */
if (drv_get_frag_size(&(sub_dev_ptr->FragSize), sub_dev_ptr->Nr) != OK){
error("%s; Failed to get fragment size!\n", drv.DriverName);
return;
}
}
if(m_ptr->COUNT != sub_dev_ptr->FragSize) {
error("Fragment size does not match user's buffer length\n");
reply(DEV_REVIVE, m_ptr->m_source, m_ptr->IO_ENDPT, EINVAL);
return;
}
/* if we are busy with something else than writing, return EBUSY */
if(sub_dev_ptr->DmaBusy && sub_dev_ptr->DmaMode != DEV_WRITE_S) {
error("Already busy with something else then writing\n");
reply(DEV_REVIVE, m_ptr->m_source, m_ptr->IO_ENDPT, EBUSY);
return;
}
sub_dev_ptr->RevivePending = TRUE;
sub_dev_ptr->ReviveProcNr = m_ptr->IO_ENDPT;
sub_dev_ptr->ReviveGrant = (cp_grant_id_t) m_ptr->ADDRESS;
sub_dev_ptr->NotifyProcNr = m_ptr->m_source;
data_from_user(sub_dev_ptr);
if(!sub_dev_ptr->DmaBusy) { /* Dma tranfer not yet started */
dprint("starting audio device\n");
get_started(sub_dev_ptr);
sub_dev_ptr->DmaMode = DEV_WRITE_S; /* Dma mode is writing */
}
}
PRIVATE void msg_read(message *m_ptr)
{
int chan; sub_dev_t *sub_dev_ptr;
special_file_t* special_file_ptr;
dprint("%s: msg_read() device %d\n", drv.DriverName, m_ptr->DEVICE);
special_file_ptr = get_special_file(m_ptr->DEVICE);
chan = special_file_ptr->read_chan;
if (chan == NO_CHANNEL) {
error("%s: No read channel specified!\n", drv.DriverName);
reply(DEV_REVIVE, m_ptr->m_source, m_ptr->IO_ENDPT, EIO);
return;
}
/* get pointer to sub device data */
sub_dev_ptr = &sub_dev[chan];
if (!sub_dev_ptr->DmaBusy) { /* get fragment size on first read */
if (drv_get_frag_size(&(sub_dev_ptr->FragSize), sub_dev_ptr->Nr) != OK){
error("%s: Could not retrieve fragment size!\n", drv.DriverName);
reply(DEV_REVIVE, m_ptr->m_source, m_ptr->IO_ENDPT, EIO);
return;
}
}
if(m_ptr->COUNT != sub_dev_ptr->FragSize) {
reply(DEV_REVIVE, m_ptr->m_source, m_ptr->IO_ENDPT, EINVAL);
error("fragment size does not match message size\n");
return;
}
/* if we are busy with something else than reading, reply EBUSY */
if(sub_dev_ptr->DmaBusy && sub_dev_ptr->DmaMode != DEV_READ_S) {
reply(DEV_REVIVE, m_ptr->m_source, m_ptr->IO_ENDPT, EBUSY);
return;
}
sub_dev_ptr->RevivePending = TRUE;
sub_dev_ptr->ReviveProcNr = m_ptr->IO_ENDPT;
sub_dev_ptr->ReviveGrant = (cp_grant_id_t) m_ptr->ADDRESS;
sub_dev_ptr->NotifyProcNr = m_ptr->m_source;
if(!sub_dev_ptr->DmaBusy) { /* Dma tranfer not yet started */
get_started(sub_dev_ptr);
sub_dev_ptr->DmaMode = DEV_READ_S; /* Dma mode is reading */
return; /* no need to get data from DMA buffer at this point */
}
/* check if data is available and possibly fill user's buffer */
data_to_user(sub_dev_ptr);
}
PRIVATE void msg_hardware(void) {
u32_t i;
dprint("%s: handling hardware message\n", drv.DriverName);
/* while we have an interrupt */
while ( drv_int_sum()) {
/* loop over all sub devices */
for ( i = 0; i < drv.NrOfSubDevices; i++) {
/* if interrupt from sub device and Dma transfer
was actually busy, take care of business */
if( drv_int(i) && sub_dev[i].DmaBusy ) {
if (sub_dev[i].DmaMode == DEV_WRITE_S) handle_int_write(i);
if (sub_dev[i].DmaMode == DEV_READ_S) handle_int_read(i);
}
}
}
/* As IRQ_REENABLE is not on in sys_irqsetpolicy, we must
* re-enable out interrupt after every interrupt.
*/
if ((sys_irqenable(&irq_hook_id)) != OK) {
error("%s: msg_hardware: Couldn't enable IRQ\n", drv.DriverName);
}
}
PRIVATE void msg_status(message *m_ptr)
{
int i;
dprint("got a status message\n");
for (i = 0; i < drv.NrOfSubDevices; i++) {
if(sub_dev[i].ReadyToRevive)
{
m_ptr->m_type = DEV_REVIVE; /* build message */
m_ptr->REP_ENDPT = sub_dev[i].ReviveProcNr;
m_ptr->REP_IO_GRANT = sub_dev[i].ReviveGrant;
m_ptr->REP_STATUS = sub_dev[i].ReviveStatus;
send(m_ptr->m_source, m_ptr); /* send the message */
/* reset variables */
sub_dev[i].ReadyToRevive = FALSE;
sub_dev[i].RevivePending = 0;
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
is_status_msg_expected = TRUE;
return; /* stop after one mess,
file system will get back for other processes */
}
}
m_ptr->m_type = DEV_NO_STATUS;
m_ptr->REP_STATUS = 0;
send(m_ptr->m_source, m_ptr); /* send DEV_NO_STATUS message */
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
is_status_msg_expected = FALSE;
}
/* handle interrupt for specified sub device; DmaMode == DEV_WRITE_S*/
PRIVATE void handle_int_write(int sub_dev_nr)
{
sub_dev_t *sub_dev_ptr;
sub_dev_ptr = &sub_dev[sub_dev_nr];
dprint("Finished playing dma[%d] ", sub_dev_ptr->DmaReadNext);
sub_dev_ptr->DmaReadNext =
(sub_dev_ptr->DmaReadNext + 1) % sub_dev_ptr->NrOfDmaFragments;
sub_dev_ptr->DmaLength -= 1;
if (sub_dev_ptr->BufLength != 0) { /* Data in extra buf, copy to Dma buf */
dprint(" buf[%d] -> dma[%d] ",
sub_dev_ptr->BufReadNext, sub_dev_ptr->DmaFillNext);
memcpy(sub_dev_ptr->DmaPtr +
sub_dev_ptr->DmaFillNext * sub_dev_ptr->FragSize,
sub_dev_ptr->ExtraBuf +
sub_dev_ptr->BufReadNext * sub_dev_ptr->FragSize,
sub_dev_ptr->FragSize);
sub_dev_ptr->BufReadNext =
(sub_dev_ptr->BufReadNext + 1) % sub_dev_ptr->NrOfExtraBuffers;
sub_dev_ptr->DmaFillNext =
(sub_dev_ptr->DmaFillNext + 1) % sub_dev_ptr->NrOfDmaFragments;
sub_dev_ptr->BufLength -= 1;
sub_dev_ptr->DmaLength += 1;
}
/* space became available, possibly copy new data from user */
data_from_user(sub_dev_ptr);
if(sub_dev_ptr->DmaLength == 0) { /* Dma buffer empty, stop Dma transfer */
sub_dev_ptr->OutOfData = TRUE; /* we're out of data */
dprint("No more work...!\n");
if (!sub_dev_ptr->Opened) {
close_sub_dev(sub_dev_ptr->Nr);
dprint("Stopping sub device %d\n", sub_dev_ptr->Nr);
return;
}
dprint("Pausing sub device %d\n",sub_dev_ptr->Nr);
drv_pause(sub_dev_ptr->Nr);
return;
}
dprint("\n");
/* confirm and reenable interrupt from this sub dev */
drv_reenable_int(sub_dev_nr);
#if 0
/* reenable irq_hook*/
if (sys_irqenable(&irq_hook_id != OK) {
error("%s Couldn't enable IRQ\n", drv.DriverName);
}
#endif
}
/* handle interrupt for specified sub device; DmaMode == DEV_READ_S */
PRIVATE void handle_int_read(int sub_dev_nr)
{
sub_dev_t *sub_dev_ptr;
sub_dev_ptr = &sub_dev[sub_dev_nr];
dprint("Device filled dma[%d]\n", sub_dev_ptr->DmaFillNext);
sub_dev_ptr->DmaLength += 1;
sub_dev_ptr->DmaFillNext =
(sub_dev_ptr->DmaFillNext + 1) % sub_dev_ptr->NrOfDmaFragments;
/* possibly copy data to user (if it is waiting for us) */
data_to_user(sub_dev_ptr);
if (sub_dev_ptr->DmaLength == sub_dev_ptr->NrOfDmaFragments) {
/* if dma buffer full */
if (sub_dev_ptr->BufLength == sub_dev_ptr->NrOfExtraBuffers) {
error("All buffers full, we have a problem.\n");
drv_stop(sub_dev_nr); /* stop the sub device */
sub_dev_ptr->DmaBusy = FALSE;
sub_dev_ptr->ReviveStatus = 0; /* no data for user,
this is a sad story */
sub_dev_ptr->ReadyToRevive = TRUE; /* wake user up */
return;
}
else { /* dma full, still room in extra buf;
copy from dma to extra buf */
dprint("dma full: going to copy buf[%d] <- dma[%d]\n",
sub_dev_ptr->BufFillNext, sub_dev_ptr->DmaReadNext);
memcpy(sub_dev_ptr->ExtraBuf +
sub_dev_ptr->BufFillNext * sub_dev_ptr->FragSize,
sub_dev_ptr->DmaPtr +
sub_dev_ptr->DmaReadNext * sub_dev_ptr->FragSize,
sub_dev_ptr->FragSize);
sub_dev_ptr->DmaLength -= 1;
sub_dev_ptr->DmaReadNext =
(sub_dev_ptr->DmaReadNext + 1) % sub_dev_ptr->NrOfDmaFragments;
sub_dev_ptr->BufFillNext =
(sub_dev_ptr->BufFillNext + 1) % sub_dev_ptr->NrOfExtraBuffers;
}
}
/* confirm interrupt, and reenable interrupt from this sub dev*/
drv_reenable_int(sub_dev_ptr->Nr);
#if 0
/* reenable irq_hook*/
if (sys_irqenable(&irq_hook_id) != OK) {
error("%s: Couldn't reenable IRQ", drv.DriverName);
}
#endif
}
PRIVATE int get_started(sub_dev_t *sub_dev_ptr) {
u32_t i;
/* enable interrupt messages from MINIX */
if ((i=sys_irqenable(&irq_hook_id)) != OK) {
error("%s: Couldn't enable IRQs: error code %u",drv.DriverName, (unsigned int) i);
return EIO;
}
/* let the lower part of the driver start the device */
if (drv_start(sub_dev_ptr->Nr, sub_dev_ptr->DmaMode) != OK) {
error("%s: Could not start device %d\n",
drv.DriverName, sub_dev_ptr->Nr);
}
sub_dev_ptr->DmaBusy = TRUE; /* Dma is busy from now on */
sub_dev_ptr->DmaReadNext = 0;
return OK;
}
PRIVATE void data_from_user(sub_dev_t *subdev)
{
int r;
message m;
if (subdev->DmaLength == subdev->NrOfDmaFragments &&
subdev->BufLength == subdev->NrOfExtraBuffers) return;/* no space */
if (!subdev->RevivePending) return; /* no new data waiting to be copied */
if (subdev->RevivePending &&
subdev->ReadyToRevive) return; /* we already got this data */
if (subdev->DmaLength < subdev->NrOfDmaFragments) { /* room in dma buf */
sys_safecopyfrom(subdev->ReviveProcNr,
(vir_bytes)subdev->ReviveGrant, 0,
(vir_bytes)subdev->DmaPtr +
subdev->DmaFillNext * subdev->FragSize,
(phys_bytes)subdev->FragSize, D);
dprint(" user -> dma[%d]\n", subdev->DmaFillNext);
subdev->DmaLength += 1;
subdev->DmaFillNext =
(subdev->DmaFillNext + 1) % subdev->NrOfDmaFragments;
} else { /* room in extra buf */
sys_safecopyfrom(subdev->ReviveProcNr,
(vir_bytes)subdev->ReviveGrant, 0,
(vir_bytes)subdev->ExtraBuf +
subdev->BufFillNext * subdev->FragSize,
(phys_bytes)subdev->FragSize, D);
dprint(" user -> buf[%d]\n", subdev->BufFillNext);
subdev->BufLength += 1;
subdev->BufFillNext =
(subdev->BufFillNext + 1) % subdev->NrOfExtraBuffers;
}
if(subdev->OutOfData) { /* if device paused (because of lack of data) */
subdev->OutOfData = FALSE;
drv_reenable_int(subdev->Nr);
/* reenable irq_hook*/
if ((sys_irqenable(&irq_hook_id)) != OK) {
error("%s: Couldn't enable IRQ", drv.DriverName);
}
drv_resume(subdev->Nr); /* resume resume the sub device */
}
subdev->ReviveStatus = subdev->FragSize;
subdev->ReadyToRevive = TRUE;
m.m_type = DEV_REVIVE; /* build message */
m.REP_ENDPT = subdev->ReviveProcNr;
m.REP_IO_GRANT = subdev->ReviveGrant;
m.REP_STATUS = subdev->ReviveStatus;
r= send(subdev->NotifyProcNr, &m); /* send the message */
if (r != OK)
{
printf("audio_fw: send to %d failed: %d\n",
subdev->NotifyProcNr, r);
}
/* reset variables */
subdev->ReadyToRevive = FALSE;
subdev->RevivePending = 0;
}
PRIVATE void data_to_user(sub_dev_t *sub_dev_ptr)
{
int r;
message m;
if (!sub_dev_ptr->RevivePending) return; /* nobody is wating for data */
if (sub_dev_ptr->ReadyToRevive) return;/* we already filled user's buffer */
if (sub_dev_ptr->BufLength == 0 && sub_dev_ptr->DmaLength == 0) return;
/* no data for user */
if(sub_dev_ptr->BufLength != 0) { /* data in extra buffer available */
sys_safecopyto(sub_dev_ptr->ReviveProcNr,
(vir_bytes)sub_dev_ptr->ReviveGrant,
0, (vir_bytes)sub_dev_ptr->ExtraBuf +
sub_dev_ptr->BufReadNext * sub_dev_ptr->FragSize,
(phys_bytes)sub_dev_ptr->FragSize, D);
dprint(" copied buf[%d] to user\n", sub_dev_ptr->BufReadNext);
/* adjust the buffer status variables */
sub_dev_ptr->BufReadNext =
(sub_dev_ptr->BufReadNext + 1) % sub_dev_ptr->NrOfExtraBuffers;
sub_dev_ptr->BufLength -= 1;
} else { /* extra buf empty, but data in dma buf*/
sys_safecopyto(
sub_dev_ptr->ReviveProcNr,
(vir_bytes)sub_dev_ptr->ReviveGrant, 0,
(vir_bytes)sub_dev_ptr->DmaPtr +
sub_dev_ptr->DmaReadNext * sub_dev_ptr->FragSize,
(phys_bytes)sub_dev_ptr->FragSize, D);
dprint(" copied dma[%d] to user\n", sub_dev_ptr->DmaReadNext);
/* adjust the buffer status variables */
sub_dev_ptr->DmaReadNext =
(sub_dev_ptr->DmaReadNext + 1) % sub_dev_ptr->NrOfDmaFragments;
sub_dev_ptr->DmaLength -= 1;
}
sub_dev_ptr->ReviveStatus = sub_dev_ptr->FragSize;
sub_dev_ptr->ReadyToRevive = TRUE;
/* drv_status will send REVIVE mess to FS*/
m.m_type = DEV_REVIVE; /* build message */
m.REP_ENDPT = sub_dev_ptr->ReviveProcNr;
m.REP_IO_GRANT = sub_dev_ptr->ReviveGrant;
m.REP_STATUS = sub_dev_ptr->ReviveStatus;
r= send(sub_dev_ptr->NotifyProcNr, &m); /* send the message */
if (r != OK)
{
printf("audio_fw: send to %d failed: %d\n",
sub_dev_ptr->NotifyProcNr, r);
}
/* reset variables */
sub_dev_ptr->ReadyToRevive = FALSE;
sub_dev_ptr->RevivePending = 0;
}
PRIVATE int init_buffers(sub_dev_t *sub_dev_ptr)
{
#if (CHIP == INTEL)
char *base;
Merge of David's ptrace branch. Summary: o Support for ptrace T_ATTACH/T_DETACH and T_SYSCALL o PM signal handling logic should now work properly, even with debuggers being present o Asynchronous PM/VFS protocol, full IPC support for senda(), and AMF_NOREPLY senda() flag DETAILS Process stop and delay call handling of PM: o Added sys_runctl() kernel call with sys_stop() and sys_resume() aliases, for PM to stop and resume a process o Added exception for sending/syscall-traced processes to sys_runctl(), and matching SIGKREADY pseudo-signal to PM o Fixed PM signal logic to deal with requests from a process after stopping it (so-called "delay calls"), using the SIGKREADY facility o Fixed various PM panics due to race conditions with delay calls versus VFS calls o Removed special PRIO_STOP priority value o Added SYS_LOCK RTS kernel flag, to stop an individual process from running while modifying its process structure Signal and debugger handling in PM: o Fixed debugger signals being dropped if a second signal arrives when the debugger has not retrieved the first one o Fixed debugger signals being sent to the debugger more than once o Fixed debugger signals unpausing process in VFS; removed PM_UNPAUSE_TR protocol message o Detached debugger signals from general signal logic and from being blocked on VFS calls, meaning that even VFS can now be traced o Fixed debugger being unable to receive more than one pending signal in one process stop o Fixed signal delivery being delayed needlessly when multiple signals are pending o Fixed wait test for tracer, which was returning for children that were not waited for o Removed second parallel pending call from PM to VFS for any process o Fixed process becoming runnable between exec() and debugger trap o Added support for notifying the debugger before the parent when a debugged child exits o Fixed debugger death causing child to remain stopped forever o Fixed consistently incorrect use of _NSIG Extensions to ptrace(): o Added T_ATTACH and T_DETACH ptrace request, to attach and detach a debugger to and from a process o Added T_SYSCALL ptrace request, to trace system calls o Added T_SETOPT ptrace request, to set trace options o Added TO_TRACEFORK trace option, to attach automatically to children of a traced process o Added TO_ALTEXEC trace option, to send SIGSTOP instead of SIGTRAP upon a successful exec() of the tracee o Extended T_GETUSER ptrace support to allow retrieving a process's priv structure o Removed T_STOP ptrace request again, as it does not help implementing debuggers properly o Added MINIX3-specific ptrace test (test42) o Added proper manual page for ptrace(2) Asynchronous PM/VFS interface: o Fixed asynchronous messages not being checked when receive() is called with an endpoint other than ANY o Added AMF_NOREPLY senda() flag, preventing such messages from satisfying the receive part of a sendrec() o Added asynsend3() that takes optional flags; asynsend() is now a #define passing in 0 as third parameter o Made PM/VFS protocol asynchronous; reintroduced tell_fs() o Made PM_BASE request/reply number range unique o Hacked in a horrible temporary workaround into RS to deal with newly revealed RS-PM-VFS race condition triangle until VFS is asynchronous System signal handling: o Fixed shutdown logic of device drivers; removed old SIGKSTOP signal o Removed is-superuser check from PM's do_procstat() (aka getsigset()) o Added sigset macros to allow system processes to deal with the full signal set, rather than just the POSIX subset Miscellaneous PM fixes: o Split do_getset into do_get and do_set, merging common code and making structure clearer o Fixed setpriority() being able to put to sleep processes using an invalid parameter, or revive zombie processes o Made find_proc() global; removed obsolete proc_from_pid() o Cleanup here and there Also included: o Fixed false-positive boot order kernel warning o Removed last traces of old NOTIFY_FROM code THINGS OF POSSIBLE INTEREST o It should now be possible to run PM at any priority, even lower than user processes o No assumptions are made about communication speed between PM and VFS, although communication must be FIFO o A debugger will now receive incoming debuggee signals at kill time only; the process may not yet be fully stopped o A first step has been made towards making the SYSTEM task preemptible
2009-09-30 11:57:22 +02:00
size_t size;
unsigned left;
u32_t i;
phys_bytes ph;
/* allocate dma buffer space */
size= sub_dev_ptr->DmaSize + 64 * 1024;
Merge of David's ptrace branch. Summary: o Support for ptrace T_ATTACH/T_DETACH and T_SYSCALL o PM signal handling logic should now work properly, even with debuggers being present o Asynchronous PM/VFS protocol, full IPC support for senda(), and AMF_NOREPLY senda() flag DETAILS Process stop and delay call handling of PM: o Added sys_runctl() kernel call with sys_stop() and sys_resume() aliases, for PM to stop and resume a process o Added exception for sending/syscall-traced processes to sys_runctl(), and matching SIGKREADY pseudo-signal to PM o Fixed PM signal logic to deal with requests from a process after stopping it (so-called "delay calls"), using the SIGKREADY facility o Fixed various PM panics due to race conditions with delay calls versus VFS calls o Removed special PRIO_STOP priority value o Added SYS_LOCK RTS kernel flag, to stop an individual process from running while modifying its process structure Signal and debugger handling in PM: o Fixed debugger signals being dropped if a second signal arrives when the debugger has not retrieved the first one o Fixed debugger signals being sent to the debugger more than once o Fixed debugger signals unpausing process in VFS; removed PM_UNPAUSE_TR protocol message o Detached debugger signals from general signal logic and from being blocked on VFS calls, meaning that even VFS can now be traced o Fixed debugger being unable to receive more than one pending signal in one process stop o Fixed signal delivery being delayed needlessly when multiple signals are pending o Fixed wait test for tracer, which was returning for children that were not waited for o Removed second parallel pending call from PM to VFS for any process o Fixed process becoming runnable between exec() and debugger trap o Added support for notifying the debugger before the parent when a debugged child exits o Fixed debugger death causing child to remain stopped forever o Fixed consistently incorrect use of _NSIG Extensions to ptrace(): o Added T_ATTACH and T_DETACH ptrace request, to attach and detach a debugger to and from a process o Added T_SYSCALL ptrace request, to trace system calls o Added T_SETOPT ptrace request, to set trace options o Added TO_TRACEFORK trace option, to attach automatically to children of a traced process o Added TO_ALTEXEC trace option, to send SIGSTOP instead of SIGTRAP upon a successful exec() of the tracee o Extended T_GETUSER ptrace support to allow retrieving a process's priv structure o Removed T_STOP ptrace request again, as it does not help implementing debuggers properly o Added MINIX3-specific ptrace test (test42) o Added proper manual page for ptrace(2) Asynchronous PM/VFS interface: o Fixed asynchronous messages not being checked when receive() is called with an endpoint other than ANY o Added AMF_NOREPLY senda() flag, preventing such messages from satisfying the receive part of a sendrec() o Added asynsend3() that takes optional flags; asynsend() is now a #define passing in 0 as third parameter o Made PM/VFS protocol asynchronous; reintroduced tell_fs() o Made PM_BASE request/reply number range unique o Hacked in a horrible temporary workaround into RS to deal with newly revealed RS-PM-VFS race condition triangle until VFS is asynchronous System signal handling: o Fixed shutdown logic of device drivers; removed old SIGKSTOP signal o Removed is-superuser check from PM's do_procstat() (aka getsigset()) o Added sigset macros to allow system processes to deal with the full signal set, rather than just the POSIX subset Miscellaneous PM fixes: o Split do_getset into do_get and do_set, merging common code and making structure clearer o Fixed setpriority() being able to put to sleep processes using an invalid parameter, or revive zombie processes o Made find_proc() global; removed obsolete proc_from_pid() o Cleanup here and there Also included: o Fixed false-positive boot order kernel warning o Removed last traces of old NOTIFY_FROM code THINGS OF POSSIBLE INTEREST o It should now be possible to run PM at any priority, even lower than user processes o No assumptions are made about communication speed between PM and VFS, although communication must be FIFO o A debugger will now receive incoming debuggee signals at kill time only; the process may not yet be fully stopped o A first step has been made towards making the SYSTEM task preemptible
2009-09-30 11:57:22 +02:00
base= alloc_contig(size, AC_ALIGN4K, &ph);
if (!base) {
error("%s: failed to allocate dma buffer for a channel\n",
drv.DriverName);
return EIO;
}
sub_dev_ptr->DmaBuf= base;
tell_dev((vir_bytes)base, size, 0, 0, 0);
/* allocate extra buffer space */
if (!(sub_dev_ptr->ExtraBuf = malloc(sub_dev_ptr->NrOfExtraBuffers *
sub_dev_ptr->DmaSize /
sub_dev_ptr->NrOfDmaFragments))) {
error("%s failed to allocate extra buffer for a channel\n",
drv.DriverName);
return EIO;
}
sub_dev_ptr->DmaPtr = sub_dev_ptr->DmaBuf;
i = sys_umap(SELF, D,
(vir_bytes) sub_dev_ptr->DmaBuf,
(phys_bytes) sizeof(sub_dev_ptr->DmaBuf),
&(sub_dev_ptr->DmaPhys));
if (i != OK) {
return EIO;
}
if ((left = dma_bytes_left(sub_dev_ptr->DmaPhys)) <
sub_dev_ptr->DmaSize) {
/* First half of buffer crosses a 64K boundary,
* can't DMA into that */
sub_dev_ptr->DmaPtr += left;
sub_dev_ptr->DmaPhys += left;
}
/* write the physical dma address and size to the device */
drv_set_dma(sub_dev_ptr->DmaPhys,
sub_dev_ptr->DmaSize, sub_dev_ptr->Nr);
return OK;
#else /* CHIP != INTEL */
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
error("%s: init_buffers() failed, CHIP != INTEL", drv.DriverName);
return EIO;
#endif /* CHIP == INTEL */
}
PRIVATE void reply(int code, int replyee, int process, int status) {
message m;
m.m_type = code; /* DEV_REVIVE */
m.REP_STATUS = status; /* result of device operation */
m.REP_ENDPT = process; /* which user made the request */
send(replyee, &m);
}
PRIVATE int io_ctl_length(int io_request) {
io_request >>= 16;
return io_request & _IOCPARM_MASK;
}
PRIVATE special_file_t* get_special_file(int minor_dev_nr) {
int i;
for(i = 0; i < drv.NrOfSpecialFiles; i++) {
if(special_file[i].minor_dev_nr == minor_dev_nr) {
return &special_file[i];
}
}
error("%s: No subdevice specified for minor device %d!\n",
drv.DriverName, minor_dev_nr);
return NULL;
}
PRIVATE void tell_dev(buf, size, pci_bus, pci_dev, pci_func)
vir_bytes buf;
size_t size;
int pci_bus;
int pci_dev;
int pci_func;
{
int r;
endpoint_t dev_e;
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
r= ds_retrieve_label_endpt("amddev", &dev_e);
if (r != OK)
{
2009-04-22 18:53:46 +02:00
#if 0
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
printf("tell_dev: ds_retrieve_label_endpt failed for 'amddev': %d\n",
r);
2009-04-22 18:53:46 +02:00
#endif
return;
}
m.m_type= IOMMU_MAP;
m.m2_i1= pci_bus;
m.m2_i2= pci_dev;
m.m2_i3= pci_func;
m.m2_l1= buf;
m.m2_l2= size;
r= sendrec(dev_e, &m);
if (r != OK)
{
printf("tell_dev: sendrec to %d failed: %d\n", dev_e, r);
return;
}
if (m.m_type != OK)
{
printf("tell_dev: dma map request failed: %d\n", m.m_type);
return;
}
}