Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
/* This file contains the multithreaded driver interface.
|
2011-11-02 17:31:38 +01:00
|
|
|
*
|
|
|
|
* Changes:
|
|
|
|
* Aug 27, 2011 created (A. Welzel)
|
|
|
|
*
|
|
|
|
* The entry points into this file are:
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
* blockdriver_mt_task: the main message loop of the driver
|
|
|
|
* blockdriver_mt_terminate: break out of the main message loop
|
|
|
|
* blockdriver_mt_sleep: put the current thread to sleep
|
|
|
|
* blockdriver_mt_wakeup: wake up a sleeping thread
|
|
|
|
* blockdriver_mt_stop: put up the current thread for termination
|
2011-11-02 17:31:38 +01:00
|
|
|
*/
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
#include <minix/blockdriver_mt.h>
|
2011-11-02 17:31:38 +01:00
|
|
|
#include <minix/mthread.h>
|
|
|
|
#include <assert.h>
|
|
|
|
|
|
|
|
#include "driver.h"
|
|
|
|
#include "mq.h"
|
|
|
|
#include "event.h"
|
|
|
|
|
|
|
|
typedef enum {
|
|
|
|
STATE_DEAD,
|
|
|
|
STATE_RUNNING,
|
|
|
|
STATE_STOPPING,
|
|
|
|
STATE_EXITED
|
|
|
|
} worker_state;
|
|
|
|
|
|
|
|
/* Structure to handle running worker threads. */
|
|
|
|
typedef struct {
|
|
|
|
thread_id_t id;
|
|
|
|
worker_state state;
|
|
|
|
mthread_thread_t mthread;
|
|
|
|
event_t queue_event;
|
|
|
|
event_t sleep_event;
|
|
|
|
} worker_t;
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PRIVATE struct blockdriver *bdtab;
|
|
|
|
PRIVATE int running = FALSE;
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
PRIVATE mthread_key_t worker_key;
|
|
|
|
|
2011-11-24 14:50:13 +01:00
|
|
|
PRIVATE worker_t worker[BLOCKDRIVER_MT_MAX_WORKERS];
|
2011-11-02 17:31:38 +01:00
|
|
|
|
2011-11-24 14:50:13 +01:00
|
|
|
PRIVATE worker_t *exited[BLOCKDRIVER_MT_MAX_WORKERS];
|
2011-11-02 17:31:38 +01:00
|
|
|
PRIVATE int num_exited = 0;
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* enqueue *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE void enqueue(worker_t *wp, const message *m_src, int ipc_status)
|
|
|
|
{
|
|
|
|
/* Enqueue a message into a worker thread's queue, and signal the thread.
|
|
|
|
* Must be called from the master thread.
|
|
|
|
*/
|
|
|
|
|
|
|
|
assert(wp->state == STATE_RUNNING || wp->state == STATE_STOPPING);
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
if (!mq_enqueue(wp->id, m_src, ipc_status))
|
|
|
|
panic("blockdriver_mt: enqueue failed (message queue full)");
|
2011-11-02 17:31:38 +01:00
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
event_fire(&wp->queue_event);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* try_dequeue *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE int try_dequeue(worker_t *wp, message *m_dst, int *ipc_status)
|
|
|
|
{
|
|
|
|
/* See if a message can be dequeued from the current worker thread's queue. If
|
|
|
|
* so, dequeue the message and return TRUE. If not, return FALSE. Must be
|
|
|
|
* called from a worker thread. Does not block.
|
|
|
|
*/
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
return mq_dequeue(wp->id, m_dst, ipc_status);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* dequeue *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE void dequeue(worker_t *wp, message *m_dst, int *ipc_status)
|
|
|
|
{
|
|
|
|
/* Dequeue a message from the current worker thread's queue. Block the current
|
|
|
|
* thread if necessary. Must be called from a worker thread. Always successful.
|
|
|
|
*/
|
|
|
|
|
|
|
|
while (!try_dequeue(wp, m_dst, ipc_status))
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
event_wait(&wp->queue_event);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* worker_thread *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE void *worker_thread(void *param)
|
|
|
|
{
|
|
|
|
/* The worker thread loop. Set up the thread-specific reference to itself and
|
|
|
|
* start looping. The loop consists of blocking dequeing and handling messages.
|
|
|
|
* After handling a message, the thread might have been stopped, so we check
|
|
|
|
* for this condition and exit if so.
|
|
|
|
*/
|
|
|
|
worker_t *wp;
|
|
|
|
message m;
|
|
|
|
int ipc_status, r;
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
|
2011-11-02 17:31:38 +01:00
|
|
|
wp = (worker_t *) param;
|
|
|
|
assert(wp != NULL);
|
|
|
|
|
|
|
|
if (mthread_setspecific(worker_key, wp))
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
panic("blockdriver_mt: could not save local thread pointer");
|
2011-11-02 17:31:38 +01:00
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
while (running) {
|
2011-11-02 17:31:38 +01:00
|
|
|
/* See if a new message is available right away. */
|
|
|
|
if (!try_dequeue(wp, &m, &ipc_status)) {
|
|
|
|
/* If not, and this thread should be stopped, stop now. */
|
|
|
|
if (wp->state == STATE_STOPPING)
|
|
|
|
break;
|
|
|
|
|
|
|
|
/* Otherwise, block waiting for a new message. */
|
|
|
|
dequeue(wp, &m, &ipc_status);
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
if (!running)
|
2011-11-02 17:31:38 +01:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Even if the thread was stopped before, a new message resumes it. */
|
|
|
|
wp->state = STATE_RUNNING;
|
|
|
|
|
|
|
|
/* Handle the request, and send a reply. */
|
2011-11-24 14:50:13 +01:00
|
|
|
r = blockdriver_handle_request(bdtab, &m, wp->id);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
blockdriver_reply(&m, ipc_status, r);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/* Clean up and terminate this thread. */
|
|
|
|
if (mthread_setspecific(worker_key, NULL))
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
panic("blockdriver_mt: could not delete local thread pointer");
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
wp->state = STATE_EXITED;
|
|
|
|
|
|
|
|
exited[num_exited++] = wp;
|
|
|
|
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* master_create_worker *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE void master_create_worker(worker_t *wp, thread_id_t id)
|
|
|
|
{
|
|
|
|
/* Start a new worker thread.
|
|
|
|
*/
|
|
|
|
int r;
|
|
|
|
|
|
|
|
wp->id = id;
|
|
|
|
wp->state = STATE_RUNNING;
|
|
|
|
|
|
|
|
/* Initialize synchronization primitives. */
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
event_init(&wp->queue_event);
|
|
|
|
event_init(&wp->sleep_event);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
r = mthread_create(&wp->mthread, NULL /*attr*/, worker_thread, (void *) wp);
|
|
|
|
|
|
|
|
if (r != 0)
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
panic("blockdriver_mt: could not start thread %d (%d)", id, r);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* master_destroy_worker *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE void master_destroy_worker(worker_t *wp)
|
|
|
|
{
|
|
|
|
/* Clean up resources used by an exited worker thread.
|
|
|
|
*/
|
|
|
|
message m;
|
|
|
|
int ipc_status;
|
|
|
|
|
|
|
|
assert(wp != NULL);
|
|
|
|
assert(wp->state == STATE_EXITED);
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
assert(!mq_dequeue(wp->id, &m, &ipc_status));
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
/* Join the thread. */
|
|
|
|
if (mthread_join(wp->mthread, NULL))
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
panic("blockdriver_mt: could not join thread %d", wp->id);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
/* Destroy resources. */
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
event_destroy(&wp->sleep_event);
|
|
|
|
event_destroy(&wp->queue_event);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
wp->state = STATE_DEAD;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* master_handle_exits *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE void master_handle_exits(void)
|
|
|
|
{
|
|
|
|
/* Destroy the remains of all exited threads.
|
|
|
|
*/
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for (i = 0; i < num_exited; i++)
|
|
|
|
master_destroy_worker(exited[i]);
|
|
|
|
|
|
|
|
num_exited = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* master_handle_request *
|
|
|
|
*===========================================================================*/
|
|
|
|
PRIVATE void master_handle_request(message *m_ptr, int ipc_status)
|
|
|
|
{
|
|
|
|
/* For real request messages, query the thread ID, start a thread if one with
|
|
|
|
* that ID is not already running, and enqueue the message in the thread's
|
|
|
|
* message queue.
|
|
|
|
*/
|
|
|
|
thread_id_t thread_id;
|
|
|
|
worker_t *wp;
|
|
|
|
int r;
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
/* If this is not a block driver request, we cannot get the minor device
|
|
|
|
* associated with it, and thus we can not tell which thread should process
|
|
|
|
* it either. In that case, the master thread has to handle it instead.
|
2011-11-02 17:31:38 +01:00
|
|
|
*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
if (!IS_BDEV_RQ(m_ptr->m_type)) {
|
|
|
|
/* Process as 'other' message. */
|
2011-11-24 14:50:13 +01:00
|
|
|
r = blockdriver_handle_request(bdtab, m_ptr, MAIN_THREAD);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
blockdriver_reply(m_ptr, ipc_status, r);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Query the thread ID. Upon failure, send the error code to the caller. */
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
r = (*bdtab->bdr_thread)(m_ptr->DEVICE, &thread_id);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
if (r != OK) {
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
blockdriver_reply(m_ptr, ipc_status, r);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Start the thread if it is not already running. */
|
2011-11-24 14:50:13 +01:00
|
|
|
assert(thread_id >= 0 && thread_id < BLOCKDRIVER_MT_MAX_WORKERS);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
wp = &worker[thread_id];
|
|
|
|
|
|
|
|
assert(wp->state != STATE_EXITED);
|
|
|
|
|
|
|
|
if (wp->state == STATE_DEAD)
|
|
|
|
master_create_worker(wp, thread_id);
|
|
|
|
|
|
|
|
/* Enqueue the message for the thread, and possibly wake it up. */
|
|
|
|
enqueue(wp, m_ptr, ipc_status);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
|
|
|
* master_init *
|
|
|
|
*===========================================================================*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PRIVATE void master_init(struct blockdriver *bdp)
|
2011-11-02 17:31:38 +01:00
|
|
|
{
|
|
|
|
/* Initialize the state of the master thread.
|
|
|
|
*/
|
|
|
|
int i;
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
assert(bdp != NULL);
|
|
|
|
assert(bdp->bdr_thread != NULL);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
mthread_init();
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
bdtab = bdp;
|
2011-11-02 17:31:38 +01:00
|
|
|
|
2011-11-24 14:50:13 +01:00
|
|
|
for (i = 0; i < BLOCKDRIVER_MT_MAX_WORKERS; i++)
|
2011-11-02 17:31:38 +01:00
|
|
|
worker[i].state = STATE_DEAD;
|
|
|
|
|
|
|
|
/* Initialize a per-thread key, where each worker thread stores its own
|
|
|
|
* reference to the worker structure.
|
|
|
|
*/
|
|
|
|
if (mthread_key_create(&worker_key, NULL))
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
panic("blockdriver_mt: error initializing worker key");
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
* blockdriver_mt_receive *
|
2011-11-02 17:31:38 +01:00
|
|
|
*===========================================================================*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PRIVATE void blockdriver_mt_receive(message *m_ptr, int *ipc_status)
|
2011-11-02 17:31:38 +01:00
|
|
|
{
|
|
|
|
/* Receive a message.
|
|
|
|
*/
|
|
|
|
int r;
|
|
|
|
|
|
|
|
r = sef_receive_status(ANY, m_ptr, ipc_status);
|
|
|
|
|
|
|
|
if (r != OK)
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
panic("blockdriver_mt: sef_receive_status() returned %d", r);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
* blockdriver_mt_task *
|
2011-11-02 17:31:38 +01:00
|
|
|
*===========================================================================*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PUBLIC void blockdriver_mt_task(struct blockdriver *driver_tab)
|
2011-11-02 17:31:38 +01:00
|
|
|
{
|
|
|
|
/* The multithreaded driver task.
|
|
|
|
*/
|
|
|
|
int ipc_status;
|
|
|
|
message mess;
|
|
|
|
|
|
|
|
/* Initialize first if necessary. */
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
if (!running) {
|
|
|
|
master_init(driver_tab);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
running = TRUE;
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/* The main message loop. */
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
while (running) {
|
2011-11-02 17:31:38 +01:00
|
|
|
/* Receive a message. */
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
blockdriver_mt_receive(&mess, &ipc_status);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
/* Dispatch the message. */
|
|
|
|
if (is_ipc_notify(ipc_status))
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
blockdriver_handle_notify(bdtab, &mess);
|
2011-11-02 17:31:38 +01:00
|
|
|
else
|
|
|
|
master_handle_request(&mess, ipc_status);
|
|
|
|
|
|
|
|
/* Let other threads run. */
|
|
|
|
mthread_yield_all();
|
|
|
|
|
|
|
|
/* Clean up any exited threads. */
|
|
|
|
if (num_exited > 0)
|
|
|
|
master_handle_exits();
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
* blockdriver_mt_terminate *
|
2011-11-02 17:31:38 +01:00
|
|
|
*===========================================================================*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PUBLIC void blockdriver_mt_terminate(void)
|
2011-11-02 17:31:38 +01:00
|
|
|
{
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
/* Instruct libblockdriver to shut down.
|
2011-11-02 17:31:38 +01:00
|
|
|
*/
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
running = FALSE;
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
* blockdriver_mt_sleep *
|
2011-11-02 17:31:38 +01:00
|
|
|
*===========================================================================*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PUBLIC void blockdriver_mt_sleep(void)
|
2011-11-02 17:31:38 +01:00
|
|
|
{
|
|
|
|
/* Let the current thread sleep until it gets woken up by the master thread.
|
|
|
|
*/
|
|
|
|
worker_t *wp;
|
|
|
|
|
|
|
|
wp = (worker_t *) mthread_getspecific(worker_key);
|
|
|
|
|
|
|
|
if (wp == NULL)
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
panic("blockdriver_mt: master thread cannot sleep");
|
2011-11-02 17:31:38 +01:00
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
event_wait(&wp->sleep_event);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
* blockdriver_mt_wakeup *
|
2011-11-02 17:31:38 +01:00
|
|
|
*===========================================================================*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PUBLIC void blockdriver_mt_wakeup(thread_id_t id)
|
2011-11-02 17:31:38 +01:00
|
|
|
{
|
|
|
|
/* Wake up a sleeping worker thread from the master thread.
|
|
|
|
*/
|
|
|
|
worker_t *wp;
|
|
|
|
|
2011-11-24 14:50:13 +01:00
|
|
|
assert(id >= 0 && id < BLOCKDRIVER_MT_MAX_WORKERS);
|
2011-11-02 17:31:38 +01:00
|
|
|
|
|
|
|
wp = &worker[id];
|
|
|
|
|
|
|
|
assert(wp->state == STATE_RUNNING || wp->state == STATE_STOPPING);
|
|
|
|
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
event_fire(&wp->sleep_event);
|
2011-11-02 17:31:38 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
/*===========================================================================*
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
* blockdriver_mt_stop *
|
2011-11-02 17:31:38 +01:00
|
|
|
*===========================================================================*/
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
PUBLIC void blockdriver_mt_stop(void)
|
2011-11-02 17:31:38 +01:00
|
|
|
{
|
|
|
|
/* Put up the current worker thread for termination. Once the current dispatch
|
|
|
|
* call has finished, and there are no more messages in the thread's message
|
|
|
|
* queue, the thread will be terminated. Any messages in the queue will undo
|
|
|
|
* the effect of this call.
|
|
|
|
*/
|
|
|
|
worker_t *wp;
|
Split block/character protocols and libdriver
This patch separates the character and block driver communication
protocols. The old character protocol remains the same, but a new
block protocol is introduced. The libdriver library is replaced by
two new libraries: libchardriver and libblockdriver. Their exposed
API, and drivers that use them, have been updated accordingly.
Together, libbdev and libblockdriver now completely abstract away
the message format used by the block protocol. As the memory driver
is both a character and a block device driver, it now implements its
own message loop.
The most important semantic change made to the block protocol is that
it is no longer possible to return both partial results and an error
for a single transfer. This simplifies the interaction between the
caller and the driver, as the I/O vector no longer needs to be copied
back. Also, drivers are now no longer supposed to decide based on the
layout of the I/O vector when a transfer should be cut short. Put
simply, transfers are now supposed to either succeed completely, or
result in an error.
After this patch, the state of the various pieces is as follows:
- block protocol: stable
- libbdev API: stable for synchronous communication
- libblockdriver API: needs slight revision (the drvlib/partition API
in particular; the threading API will also change shortly)
- character protocol: needs cleanup
- libchardriver API: needs cleanup accordingly
- driver restarts: largely unsupported until endpoint changes are
reintroduced
As a side effect, this patch eliminates several bugs, hacks, and gcc
-Wall and -W warnings all over the place. It probably introduces a
few new ones, too.
Update warning: this patch changes the protocol between MFS and disk
drivers, so in order to use old/new images, the MFS from the ramdisk
must be used to mount all file systems.
2011-11-22 13:27:53 +01:00
|
|
|
|
2011-11-02 17:31:38 +01:00
|
|
|
wp = (worker_t *) mthread_getspecific(worker_key);
|
|
|
|
|
|
|
|
assert(wp != NULL);
|
|
|
|
assert(wp->state == STATE_RUNNING || wp->state == STATE_STOPPING);
|
|
|
|
|
|
|
|
wp->state = STATE_STOPPING;
|
|
|
|
}
|