d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
1f9c1db5f5
freebsd-dev/sys/kern/subr_msgbuf.c

418 lines
11 KiB
C
Raw Normal View History

/* -> /*- for copyright notices, minor format tweaks as necessary
2005-01-06 23:35:40 +00:00
/*-
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
* Copyright (c) 2003 Ian Dowse. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
* $FreeBSD$
*/
/*
* Generic message buffer support routines.
*/
#include <sys/param.h>
#include <sys/systm.h>
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
#include <sys/lock.h>
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
#include <sys/kernel.h>
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
#include <sys/mutex.h>
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
#include <sys/msgbuf.h>
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
#include <sys/sysctl.h>
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
/*
* Maximum number conversion buffer length: uintmax_t in base 2, plus <>
* around the priority, and a terminating NUL.
*/
#define MAXPRIBUF (sizeof(intmax_t) * NBBY + 3)
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
/* Read/write sequence numbers are modulo a multiple of the buffer size. */
#define SEQMOD(size) ((size) * 16)
static u_int msgbuf_cksum(struct msgbuf *mbp);
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
/*
- Clean up timestamps in msgbuf code. The timestamps should now be inserted after the priority token thus cleaning up the output. - Remove the needless double internal do_add_char function. - Resolve a possible deadlock if interrupts are disabled and getnanotime is called Reviewed by: bde kmacy, avg, sbruno (various versions) Approved by: cperciva MFC after: 2 weeks
2012-03-19 00:36:32 +00:00
* Timestamps in msgbuf are useful when trying to diagnose when core dumps
* or other actions occured.
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
*/
static int msgbuf_show_timestamp = 0;
Pull in r267961 and r267973 again. Fix for issues reported will follow.
2014-06-28 03:56:17 +00:00
SYSCTL_INT(_kern, OID_AUTO, msgbuf_show_timestamp, CTLFLAG_RWTUN,
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
&msgbuf_show_timestamp, 0, "Show timestamp in msgbuf");
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
/*
* Initialize a message buffer of the specified size at the specified
* location. This also zeros the buffer area.
*/
void
msgbuf_init(struct msgbuf *mbp, void *ptr, int size)
{
mbp->msg_ptr = ptr;
mbp->msg_size = size;
mbp->msg_seqmod = SEQMOD(size);
msgbuf_clear(mbp);
mbp->msg_magic = MSG_MAGIC;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mbp->msg_lastpri = -1;
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
mbp->msg_flags = 0;
Fix a bug introduced in revision 222537. In msgbuf_reinit() and msgbuf_init(), we weren't initializing the mutex. Depending on the contents of memory, the LO_INITIALIZED flag might be set on the mutex (either due to a warm reboot, and the message buffer remaining in place, or due to garbage in memory) and in that case, with INVARIANTS turned on, we would trigger an assertion that the mutex had already been initialized. Fix this by bzeroing the message buffer mutex for the _init() and _reinit() paths. Reported by: mdf
2011-05-31 22:39:32 +00:00
bzero(&mbp->msg_lock, sizeof(mbp->msg_lock));
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_init(&mbp->msg_lock, "msgbuf", NULL, MTX_SPIN);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
}
/*
* Reinitialize a message buffer, retaining its previous contents if
* the size and checksum are correct. If the old contents cannot be
* recovered, the message buffer is cleared.
*/
void
msgbuf_reinit(struct msgbuf *mbp, void *ptr, int size)
{
u_int cksum;
if (mbp->msg_magic != MSG_MAGIC || mbp->msg_size != size) {
msgbuf_init(mbp, ptr, size);
return;
}
mbp->msg_seqmod = SEQMOD(size);
mbp->msg_wseq = MSGBUF_SEQNORM(mbp, mbp->msg_wseq);
mbp->msg_rseq = MSGBUF_SEQNORM(mbp, mbp->msg_rseq);
mbp->msg_ptr = ptr;
cksum = msgbuf_cksum(mbp);
if (cksum != mbp->msg_cksum) {
Put the message about msgbuf cksum mismatch under bootverbose and tell people what the consequence is.
2003-09-05 11:12:00 +00:00
if (bootverbose) {
printf("msgbuf cksum mismatch (read %x, calc %x)\n",
mbp->msg_cksum, cksum);
printf("Old msgbuf not recovered\n");
}
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
msgbuf_clear(mbp);
}
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mbp->msg_lastpri = -1;
/* Assume that the old message buffer didn't end in a newline. */
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
mbp->msg_flags |= MSGBUF_NEEDNL;
Fix a bug introduced in revision 222537. In msgbuf_reinit() and msgbuf_init(), we weren't initializing the mutex. Depending on the contents of memory, the LO_INITIALIZED flag might be set on the mutex (either due to a warm reboot, and the message buffer remaining in place, or due to garbage in memory) and in that case, with INVARIANTS turned on, we would trigger an assertion that the mutex had already been initialized. Fix this by bzeroing the message buffer mutex for the _init() and _reinit() paths. Reported by: mdf
2011-05-31 22:39:32 +00:00
bzero(&mbp->msg_lock, sizeof(mbp->msg_lock));
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_init(&mbp->msg_lock, "msgbuf", NULL, MTX_SPIN);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
}
/*
* Clear the message buffer.
*/
void
msgbuf_clear(struct msgbuf *mbp)
{
bzero(mbp->msg_ptr, mbp->msg_size);
mbp->msg_wseq = 0;
mbp->msg_rseq = 0;
mbp->msg_cksum = 0;
}
/*
* Get a count of the number of unread characters in the message buffer.
*/
int
msgbuf_getcount(struct msgbuf *mbp)
{
u_int len;
len = MSGBUF_SEQSUB(mbp, mbp->msg_wseq, mbp->msg_rseq);
if (len > mbp->msg_size)
len = mbp->msg_size;
return (len);
}
/*
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
* Add a character into the message buffer, and update the checksum and
* sequence number.
*
* The caller should hold the message buffer spinlock.
*/
- Clean up timestamps in msgbuf code. The timestamps should now be inserted after the priority token thus cleaning up the output. - Remove the needless double internal do_add_char function. - Resolve a possible deadlock if interrupts are disabled and getnanotime is called Reviewed by: bde kmacy, avg, sbruno (various versions) Approved by: cperciva MFC after: 2 weeks
2012-03-19 00:36:32 +00:00
static void
msgbuf_do_addchar(struct msgbuf * const mbp, u_int * const seq, const int c)
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
{
u_int pos;
/* Make sure we properly wrap the sequence number. */
pos = MSGBUF_SEQ_TO_POS(mbp, *seq);
- Clean up timestamps in msgbuf code. The timestamps should now be inserted after the priority token thus cleaning up the output. - Remove the needless double internal do_add_char function. - Resolve a possible deadlock if interrupts are disabled and getnanotime is called Reviewed by: bde kmacy, avg, sbruno (various versions) Approved by: cperciva MFC after: 2 weeks
2012-03-19 00:36:32 +00:00
mbp->msg_cksum += (u_int)(u_char)c -
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
(u_int)(u_char)mbp->msg_ptr[pos];
mbp->msg_ptr[pos] = c;
*seq = MSGBUF_SEQNORM(mbp, *seq + 1);
}
/*
* Append a character to a message buffer.
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
*/
void
msgbuf_addchar(struct msgbuf *mbp, int c)
{
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_lock_spin(&mbp->msg_lock);
msgbuf_do_addchar(mbp, &mbp->msg_wseq, c);
mtx_unlock_spin(&mbp->msg_lock);
}
/*
* Append a NUL-terminated string with a priority to a message buffer.
* Filter carriage returns if the caller requests it.
*
* XXX The carriage return filtering behavior is present in the
* msglogchar() API, however testing has shown that we don't seem to send
* carriage returns down this path. So do we still need it?
*/
void
msgbuf_addstr(struct msgbuf *mbp, int pri, char *str, int filter_cr)
{
u_int seq;
size_t len, prefix_len;
char prefix[MAXPRIBUF];
- Clean up timestamps in msgbuf code. The timestamps should now be inserted after the priority token thus cleaning up the output. - Remove the needless double internal do_add_char function. - Resolve a possible deadlock if interrupts are disabled and getnanotime is called Reviewed by: bde kmacy, avg, sbruno (various versions) Approved by: cperciva MFC after: 2 weeks
2012-03-19 00:36:32 +00:00
char buf[32];
int nl, i, j, needtime;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
len = strlen(str);
prefix_len = 0;
nl = 0;
/* If we have a zero-length string, no need to do anything. */
if (len == 0)
return;
mtx_lock_spin(&mbp->msg_lock);
/*
* If this is true, we may need to insert a new priority sequence,
* so prepare the prefix.
*/
if (pri != -1)
prefix_len = sprintf(prefix, "<%d>", pri);
/*
* Starting write sequence number.
*/
seq = mbp->msg_wseq;
/*
* Whenever there is a change in priority, we have to insert a
* newline, and a priority prefix if the priority is not -1. Here
* we detect whether there was a priority change, and whether we
* did not end with a newline. If that is the case, we need to
* insert a newline before this string.
*/
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
if (mbp->msg_lastpri != pri && (mbp->msg_flags & MSGBUF_NEEDNL) != 0) {
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
msgbuf_do_addchar(mbp, &seq, '\n');
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
mbp->msg_flags &= ~MSGBUF_NEEDNL;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
}
- Clean up timestamps in msgbuf code. The timestamps should now be inserted after the priority token thus cleaning up the output. - Remove the needless double internal do_add_char function. - Resolve a possible deadlock if interrupts are disabled and getnanotime is called Reviewed by: bde kmacy, avg, sbruno (various versions) Approved by: cperciva MFC after: 2 weeks
2012-03-19 00:36:32 +00:00
needtime = 1;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
for (i = 0; i < len; i++) {
/*
* If we just had a newline, and the priority is not -1
* (and therefore prefix_len != 0), then we need a priority
* prefix for this line.
*/
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
if ((mbp->msg_flags & MSGBUF_NEEDNL) == 0 && prefix_len != 0) {
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
int j;
for (j = 0; j < prefix_len; j++)
msgbuf_do_addchar(mbp, &seq, prefix[j]);
}
- Clean up timestamps in msgbuf code. The timestamps should now be inserted after the priority token thus cleaning up the output. - Remove the needless double internal do_add_char function. - Resolve a possible deadlock if interrupts are disabled and getnanotime is called Reviewed by: bde kmacy, avg, sbruno (various versions) Approved by: cperciva MFC after: 2 weeks
2012-03-19 00:36:32 +00:00
if (msgbuf_show_timestamp && needtime == 1 &&
(mbp->msg_flags & MSGBUF_NEEDNL) == 0) {
snprintf(buf, sizeof(buf), "[%jd] ",
(intmax_t)time_uptime);
for (j = 0; buf[j] != '\0'; j++)
msgbuf_do_addchar(mbp, &seq, buf[j]);
needtime = 0;
}
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
/*
* Don't copy carriage returns if the caller requested
* filtering.
*
* XXX This matches the behavior of msglogchar(), but is it
* necessary? Testing has shown that we don't seem to get
* carriage returns here.
*/
if ((filter_cr != 0) && (str[i] == '\r'))
continue;
/*
* Clear this flag if we see a newline. This affects whether
* we need to insert a new prefix or insert a newline later.
*/
if (str[i] == '\n')
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
mbp->msg_flags &= ~MSGBUF_NEEDNL;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
else
Add a timestamp to the msgbuf output in order to determine when when messages were printed. This can be enabled with the kern.msgbuf_show_timestamp sysctl PR: kern/161553 Reviewed by: avg Submitted by: Arnaud Lacombe <lacombar@gmail.com> Approved by: cperciva MFC after: 1 month
2012-02-16 05:11:35 +00:00
mbp->msg_flags |= MSGBUF_NEEDNL;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
msgbuf_do_addchar(mbp, &seq, str[i]);
}
/*
* Update the write sequence number for the actual number of
* characters we put in the message buffer. (Depends on whether
* carriage returns are filtered.)
*/
mbp->msg_wseq = seq;
/*
* Set the last priority.
*/
mbp->msg_lastpri = pri;
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
}
/*
* Read and mark as read a character from a message buffer.
* Returns the character, or -1 if no characters are available.
*/
int
msgbuf_getchar(struct msgbuf *mbp)
{
u_int len, wseq;
int c;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_lock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
wseq = mbp->msg_wseq;
len = MSGBUF_SEQSUB(mbp, wseq, mbp->msg_rseq);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
if (len == 0) {
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
return (-1);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
}
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
if (len > mbp->msg_size)
mbp->msg_rseq = MSGBUF_SEQNORM(mbp, wseq - mbp->msg_size);
c = (u_char)mbp->msg_ptr[MSGBUF_SEQ_TO_POS(mbp, mbp->msg_rseq)];
mbp->msg_rseq = MSGBUF_SEQNORM(mbp, mbp->msg_rseq + 1);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
return (c);
}
/*
* Read and mark as read a number of characters from a message buffer.
* Returns the number of characters that were placed in `buf'.
*/
int
msgbuf_getbytes(struct msgbuf *mbp, char *buf, int buflen)
{
u_int len, pos, wseq;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_lock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
wseq = mbp->msg_wseq;
len = MSGBUF_SEQSUB(mbp, wseq, mbp->msg_rseq);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
if (len == 0) {
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
return (0);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
}
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
if (len > mbp->msg_size) {
mbp->msg_rseq = MSGBUF_SEQNORM(mbp, wseq - mbp->msg_size);
len = mbp->msg_size;
}
pos = MSGBUF_SEQ_TO_POS(mbp, mbp->msg_rseq);
len = min(len, mbp->msg_size - pos);
len = min(len, (u_int)buflen);
bcopy(&mbp->msg_ptr[pos], buf, len);
mbp->msg_rseq = MSGBUF_SEQNORM(mbp, mbp->msg_rseq + len);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
return (len);
}
/*
* Peek at the full contents of a message buffer without marking any
* data as read. `seqp' should point to an unsigned integer that
* msgbuf_peekbytes() can use to retain state between calls so that
* the whole message buffer can be read in multiple short reads.
* To initialise this variable to the start of the message buffer,
* call msgbuf_peekbytes() with a NULL `buf' parameter.
*
* Returns the number of characters that were placed in `buf'.
*/
int
msgbuf_peekbytes(struct msgbuf *mbp, char *buf, int buflen, u_int *seqp)
{
u_int len, pos, wseq;
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_lock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
if (buf == NULL) {
/* Just initialise *seqp. */
*seqp = MSGBUF_SEQNORM(mbp, mbp->msg_wseq - mbp->msg_size);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
return (0);
}
wseq = mbp->msg_wseq;
len = MSGBUF_SEQSUB(mbp, wseq, *seqp);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
if (len == 0) {
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
return (0);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
}
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
if (len > mbp->msg_size) {
*seqp = MSGBUF_SEQNORM(mbp, wseq - mbp->msg_size);
len = mbp->msg_size;
}
pos = MSGBUF_SEQ_TO_POS(mbp, *seqp);
len = min(len, mbp->msg_size - pos);
len = min(len, (u_int)buflen);
bcopy(&mbp->msg_ptr[MSGBUF_SEQ_TO_POS(mbp, *seqp)], buf, len);
*seqp = MSGBUF_SEQNORM(mbp, *seqp + len);
Fix apparent garbage in the message buffer. While we have had a fix in place (options PRINTF_BUFR_SIZE=128) to fix scrambled console output, the message buffer and syslog were still getting log messages one character at a time. While all of the characters still made it into the log (courtesy of atomic operations), they were often interleaved when there were multiple threads writing to the buffer at the same time. This fixes message buffer accesses to use buffering logic as well, so that strings that are less than PRINTF_BUFR_SIZE will be put into the message buffer atomically. So now dmesg output should look the same as console output. subr_msgbuf.c: Convert most message buffer calls to use a new spin lock instead of atomic variables in some places. Add a new routine, msgbuf_addstr(), that adds a NUL-terminated string to a message buffer. This takes a priority argument, which allows us to eliminate some races (at least in the the string at a time case) that are present in the implementation of msglogchar(). (dangling and lastpri are static variables, and are subject to races when multiple callers are present.) msgbuf_addstr() also allows the caller to request that carriage returns be stripped out of the string. This matches the behavior of msglogchar(), but in testing so far it doesn't appear that any newlines are being stripped out. So the carriage return removal functionality may be a candidate for removal later on if further analysis shows that it isn't necessary. subr_prf.c: Add a new msglogstr() routine that calls msgbuf_logstr(). Rename putcons() to putbuf(). This now handles buffered output to the message log as well as the console. Also, remove the logic in putcons() (now putbuf()) that added a carriage return before a newline. The console path was the only path that needed it, and cnputc() (called by cnputs()) already adds a carriage return. So this duplication resulted in kernel-generated console output lines ending in '\r''\r''\n'. Refactor putchar() to handle the new buffering scheme. Add buffering to log(). Change log_console() to use msglogstr() instead of msglogchar(). Don't add extra newlines by default in log_console(). Hide that behavior behind a tunable/sysctl (kern.log_console_add_linefeed) for those who would like the old behavior. The old behavior led to the insertion of extra newlines for log output for programs that print out a string, and then a trailing newline on a separate write. (This is visible with dmesg -a.) msgbuf.h: Add a prototype for msgbuf_addstr(). Add three new fields to struct msgbuf, msg_needsnl, msg_lastpri and msg_lock. The first two are needed for log message functionality previously handled by msglogchar(). (Which is still active if buffering isn't enabled.) Include sys/lock.h and sys/mutex.h for the new mutex. Reviewed by: gibbs
2011-05-31 17:29:58 +00:00
mtx_unlock_spin(&mbp->msg_lock);
Replace the code for reading and writing the kernel message buffer with a new implementation that has a mostly reentrant "addchar" routine, supports multiple message buffers in the kernel, and hides the implementation details from callers. The new code uses a kind of sequence number to represend the current read and write positions in the buffer. This approach (suggested mainly by bde) permits the read and write pointers to be maintained separately, which reduces the number of atomic operations that are required. The "mostly reentrant" above refers to the way that while it is now always safe to have any number of concurrent writers, readers could see the message buffer after a writer has advanced the pointers but before it has witten the new character. Discussed on: freebsd-arch
2003-06-22 02:18:31 +00:00
return (len);
}
/*
* Compute the checksum for the complete message buffer contents.
*/
static u_int
msgbuf_cksum(struct msgbuf *mbp)
{
u_int i, sum;
sum = 0;
for (i = 0; i < mbp->msg_size; i++)
sum += (u_char)mbp->msg_ptr[i];
return (sum);
}
/*
* Copy from one message buffer to another.
*/
void
msgbuf_copy(struct msgbuf *src, struct msgbuf *dst)
{
int c;
while ((c = msgbuf_getchar(src)) >= 0)
msgbuf_addchar(dst, c);
}
Reference in New Issue Copy Permalink