d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
560a8e6ccc
freebsd-skq/sys/ddb/db_ps.c

509 lines
13 KiB
C
Raw Normal View History

ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
/*-
sys: further adoption of SPDX licensing ID tags. Mainly focus on files that use BSD 3-Clause license. The Software Package Data Exchange (SPDX) group provides a specification to make it easier for automated tools to detect and summarize well known opensource licenses. We are gradually adopting the specification, noting that the tags are considered only advisory and do not, in any way, superceed or replace the license texts. Special thanks to Wind River for providing access to "The Duke of Highlander" tool: an older (2014) run over FreeBSD tree was useful as a starting point.
2017-11-20 19:43:44 +00:00
* SPDX-License-Identifier: BSD-3-Clause
*
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
* Copyright (c) 1993 The Regents of the University of California.
* 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.
Renumber copyright clause 4 Renumber cluase 4 to 3, per what everybody else did when BSD granted them permission to remove clause 3. My insistance on keeping the same numbering for legal reasons is too pedantic, so give up on that point. Submitted by: Jan Schaumann <jschauma@stevens.edu> Pull Request: https://github.com/freebsd/freebsd/pull/96
2017-02-28 23:42:47 +00:00
* 3. Neither the name of the University nor the names of its contributors
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS 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 REGENTS 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.
*/
Use __FBSDID().
2003-06-10 22:09:23 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
Fix KSTACK_PAGES issue when the default value was changed in KERNCONF If KSTACK_PAGES was changed to anything alse than the default, the value from param.h was taken instead in some places and the value from KENRCONF in some others. This resulted in inconsistency which caused corruption in SMP envorinment. Ensure all places where KSTACK_PAGES are used the opt_kstack_pages.h is included. The file opt_kstack_pages.h could not be included in param.h because was breaking the toolchain compilation. Reviewed by: kib Obtained from: Semihalf Sponsored by: The FreeBSD Foundation Differential Revision: https://reviews.freebsd.org/D3094
2015-07-16 10:46:52 +00:00
#include "opt_kstack_pages.h"
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
#include <sys/param.h>
Sort includes.
2006-04-27 22:09:18 +00:00
#include <sys/cons.h>
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#include <sys/jail.h>
#include <sys/kdb.h>
Add td_swinvoltick to track last involuntary context switch Expose in DDB via "show thread." Reviewed by: markj Sponsored by: EMC / Isilon Storage Division
2016-03-25 19:35:29 +00:00
#include <sys/kernel.h>
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
#include <sys/proc.h>
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#include <sys/sysent.h>
Sort includes.
2006-04-27 22:09:18 +00:00
#include <sys/systm.h>
Add 'findstack' ddb command to search either the thread kernel stack or cached stack containing the specified kernel virtual address. Discussed with: pho MFC after: 1 week
2011-12-16 11:44:20 +00:00
#include <sys/_kstack_cache.h>
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
#include <vm/vm.h>
#include <vm/vm_param.h>
#include <vm/pmap.h>
Extend show proc with reaper, sigparent, and vmspace information I have regularly needed the last couple of months. Sponsored by: iXsystems, Inc.
2018-05-25 13:59:48 +00:00
#include <vm/vm_map.h>
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
Reduced and cleaned up #includes.
1995-12-10 19:08:32 +00:00
#include <ddb/ddb.h>
Whitespace nits.
2003-06-06 17:19:27 +00:00
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
static void dumpthread(volatile struct proc *p, volatile struct thread *td,
int all);
enable dynamic addition of "show all" commands MFC after: 3 weeks
2008-03-25 20:36:32 +00:00
/*
* At least one non-optional show-command must be implemented using
* DB_SHOW_ALL_COMMAND() so that db_show_all_cmd_set gets created.
* Here is one.
*/
DB_SHOW_ALL_COMMAND(procs, db_procs_cmd)
{
db_ps(addr, have_addr, count, modif);
}
Reduced and cleaned up #includes.
1995-12-10 19:08:32 +00:00
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
/*
* Layout:
* - column counts
* - header
* - single-threaded process
* - multi-threaded process
* - thread in a MT process
*
* 1 2 3 4 5 6 7
* 1234567890123456789012345678901234567890123456789012345678901234567890
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
* pid ppid pgrp uid state wmesg wchan cmd
* <pid> <ppi> <pgi> <uid> <stat> <wmesg> <wchan > <name>
Fix two nits in the ps header that offset each other making them largely unnoticable.
2006-08-01 22:30:55 +00:00
* <pid> <ppi> <pgi> <uid> <stat> (threaded) <command>
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
* <tid > <stat> <wmesg> <wchan > <name>
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
*
* For machines with 64-bit pointers, we expand the wchan field 8 more
* characters.
*/
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
void
ddb: finish converting boolean values. The replacement started at r283088 was necessarily incomplete without replacing boolean_t with bool. This also involved cleaning some type mismatches and ansifying old C function declarations. Pointed out by: bde Discussed with: bde, ian, jhb
2015-05-21 15:16:18 +00:00
db_ps(db_expr_t addr, bool hasaddr, db_expr_t count, char *modif)
Change db_ps() to have the correct number and type of args for a ddb command.
1995-08-27 02:40:39 +00:00
{
Make db_ps tolerant of uninitialized struct proc fields. Improved 'more' function to work like you'd expect. Other misc cleanup.
1995-08-20 05:25:11 +00:00
volatile struct proc *p, *pp;
Make the flag field in the ps output one char wider to account for recent growth in the number of flags used. Also, if a thread is blocked on a mutex, print the mutex name in the wait channel column.
2001-10-20 03:22:23 +00:00
volatile struct thread *td;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
struct ucred *cred;
struct pgrp *pgrp;
char state[9];
Simplify the pager support in DDB. Allowing different db commands to install custom pager functions didn't actually happen in practice (they all just used the simple pager and passed in a local quit pointer). So, just hardcode the simple pager as the only pager and make it set a global db_pager_quit flag that db commands can check when the user hits 'q' (or a suitable variant) at the pager prompt. Also, now that it's easy to do so, enable paging by default for all ddb commands. Any command that wishes to honor the quit flag can do so by checking db_pager_quit. Note that the pager can also be effectively disabled by setting $lines to 0. Other fixes: - 'show idt' on i386 and pc98 now actually checks the quit flag and terminates early. - 'show intr' now actually checks the quit flag and terminates early.
2006-07-12 21:22:44 +00:00
int np, rflag, sflag, dflag, lflag, wflag;
Make db_ps tolerant of uninitialized struct proc fields. Improved 'more' function to work like you'd expect. Other misc cleanup.
1995-08-20 05:25:11 +00:00
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
np = nprocs;
Make db_ps tolerant of uninitialized struct proc fields. Improved 'more' function to work like you'd expect. Other misc cleanup.
1995-08-20 05:25:11 +00:00
Use macro API for <sys/queue.h>
2000-12-30 22:06:19 +00:00
if (!LIST_EMPTY(&allproc))
p = LIST_FIRST(&allproc);
Make db_ps tolerant of uninitialized struct proc fields. Improved 'more' function to work like you'd expect. Other misc cleanup.
1995-08-20 05:25:11 +00:00
else
p = &proc0;
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
Use __LP64__ rather than the PTR64 hack. Suggested by: ru
2006-05-11 21:59:55 +00:00
#ifdef __LP64__
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
db_printf(" pid ppid pgrp uid state wmesg wchan cmd\n");
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#else
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
db_printf(" pid ppid pgrp uid state wmesg wchan cmd\n");
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#endif
Simplify the pager support in DDB. Allowing different db commands to install custom pager functions didn't actually happen in practice (they all just used the simple pager and passed in a local quit pointer). So, just hardcode the simple pager as the only pager and make it set a global db_pager_quit flag that db commands can check when the user hits 'q' (or a suitable variant) at the pager prompt. Also, now that it's easy to do so, enable paging by default for all ddb commands. Any command that wishes to honor the quit flag can do so by checking db_pager_quit. Note that the pager can also be effectively disabled by setting $lines to 0. Other fixes: - 'show idt' on i386 and pc98 now actually checks the quit flag and terminates early. - 'show intr' now actually checks the quit flag and terminates early.
2006-07-12 21:22:44 +00:00
while (--np >= 0 && !db_pager_quit) {
A small bit of defensive programming in case the panic is during process exit and cleanup. the 'ps' command assumes that there are always 'nproc' processes on the lists and will walk off the end without checking if not, causing ddb to trap during the 'ps' command.
1996-06-15 07:08:02 +00:00
if (p == NULL) {
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf("oops, ran out of processes early!\n");
A small bit of defensive programming in case the panic is during process exit and cleanup. the 'ps' command assumes that there are always 'nproc' processes on the lists and will walk off the end without checking if not, causing ddb to trap during the 'ps' command.
1996-06-15 07:08:02 +00:00
break;
}
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
pp = p->p_pptr;
Make db_ps tolerant of uninitialized struct proc fields. Improved 'more' function to work like you'd expect. Other misc cleanup.
1995-08-20 05:25:11 +00:00
if (pp == NULL)
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
pp = p;
Make db_ps tolerant of uninitialized struct proc fields. Improved 'more' function to work like you'd expect. Other misc cleanup.
1995-08-20 05:25:11 +00:00
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
cred = p->p_ucred;
pgrp = p->p_pgrp;
Some cosmetic tweaks: - Right justify 'pid' label. - Move the uid column to the right 2 columns so that the 3 process id columns (pid, ppid, pgrp) are grouped together. - Expand the uid column to 5 chars. - Don't indent the tid for multithreaded processes. Requested by: bde (1, 2, 4)
2006-08-01 15:29:46 +00:00
db_printf("%5d %5d %5d %5d ", p->p_pid, pp->p_pid,
pgrp != NULL ? pgrp->pg_id : 0,
cred != NULL ? cred->cr_ruid : 0);
Part 1 of KSE-III The ability to schedule multiple threads per process (one one cpu) by making ALL system calls optionally asynchronous. to come: ia64 and power-pc patches, patches for gdb, test program (in tools) Reviewed by: Almost everyone who counts (at various times, peter, jhb, matt, alfred, mini, bernd, and a cast of thousands) NOTE: this is still Beta code, and contains lots of debugging stuff. expect slight instability in signals..
2002-06-29 17:26:22 +00:00
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
/* Determine our primary process state. */
A whitespace fix. Submitted by: bde
2006-04-27 22:02:27 +00:00
switch (p->p_state) {
Part 1 of KSE-III The ability to schedule multiple threads per process (one one cpu) by making ALL system calls optionally asynchronous. to come: ia64 and power-pc patches, patches for gdb, test program (in tools) Reviewed by: Almost everyone who counts (at various times, peter, jhb, matt, alfred, mini, bernd, and a cast of thousands) NOTE: this is still Beta code, and contains lots of debugging stuff. expect slight instability in signals..
2002-06-29 17:26:22 +00:00
case PRS_NORMAL:
if (P_SHOULDSTOP(p))
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
state[0] = 'T';
else {
/*
* One of D, L, R, S, W. For a
* multithreaded process we will use
* the state of the thread with the
* highest precedence. The
* precendence order from high to low
* is R, L, D, S, W. If no thread is
* in a sane state we use '?' for our
* primary state.
*/
rflag = sflag = dflag = lflag = wflag = 0;
FOREACH_THREAD_IN_PROC(p, td) {
if (td->td_state == TDS_RUNNING ||
td->td_state == TDS_RUNQ ||
td->td_state == TDS_CAN_RUN)
rflag++;
if (TD_ON_LOCK(td))
lflag++;
if (TD_IS_SLEEPING(td)) {
Replace several instances of 'if (!a & b)' with 'if (!(a &b))' in order to silence newer GCC versions.
2010-01-08 15:44:49 +00:00
if (!(td->td_flags & TDF_SINTR))
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
dflag++;
else
sflag++;
}
if (TD_AWAITING_INTR(td))
wflag++;
}
if (rflag)
state[0] = 'R';
else if (lflag)
state[0] = 'L';
else if (dflag)
state[0] = 'D';
else if (sflag)
state[0] = 'S';
else if (wflag)
state[0] = 'W';
else
ddb: space/tab fixes. No functional change. MFC after: 3 days
2014-10-11 20:25:19 +00:00
state[0] = '?';
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
}
Part 1 of KSE-III The ability to schedule multiple threads per process (one one cpu) by making ALL system calls optionally asynchronous. to come: ia64 and power-pc patches, patches for gdb, test program (in tools) Reviewed by: Almost everyone who counts (at various times, peter, jhb, matt, alfred, mini, bernd, and a cast of thousands) NOTE: this is still Beta code, and contains lots of debugging stuff. expect slight instability in signals..
2002-06-29 17:26:22 +00:00
break;
case PRS_NEW:
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
state[0] = 'N';
Part 1 of KSE-III The ability to schedule multiple threads per process (one one cpu) by making ALL system calls optionally asynchronous. to come: ia64 and power-pc patches, patches for gdb, test program (in tools) Reviewed by: Almost everyone who counts (at various times, peter, jhb, matt, alfred, mini, bernd, and a cast of thousands) NOTE: this is still Beta code, and contains lots of debugging stuff. expect slight instability in signals..
2002-06-29 17:26:22 +00:00
break;
case PRS_ZOMBIE:
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
state[0] = 'Z';
Part 1 of KSE-III The ability to schedule multiple threads per process (one one cpu) by making ALL system calls optionally asynchronous. to come: ia64 and power-pc patches, patches for gdb, test program (in tools) Reviewed by: Almost everyone who counts (at various times, peter, jhb, matt, alfred, mini, bernd, and a cast of thousands) NOTE: this is still Beta code, and contains lots of debugging stuff. expect slight instability in signals..
2002-06-29 17:26:22 +00:00
break;
default:
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
state[0] = 'U';
Part 1 of KSE-III The ability to schedule multiple threads per process (one one cpu) by making ALL system calls optionally asynchronous. to come: ia64 and power-pc patches, patches for gdb, test program (in tools) Reviewed by: Almost everyone who counts (at various times, peter, jhb, matt, alfred, mini, bernd, and a cast of thousands) NOTE: this is still Beta code, and contains lots of debugging stuff. expect slight instability in signals..
2002-06-29 17:26:22 +00:00
break;
}
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
state[1] = '\0';
/* Additional process state flags. */
Replace several instances of 'if (!a & b)' with 'if (!(a &b))' in order to silence newer GCC versions.
2010-01-08 15:44:49 +00:00
if (!(p->p_flag & P_INMEM))
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
strlcat(state, "W", sizeof(state));
if (p->p_flag & P_TRACED)
strlcat(state, "X", sizeof(state));
if (p->p_flag & P_WEXIT && p->p_state != PRS_ZOMBIE)
strlcat(state, "E", sizeof(state));
if (p->p_flag & P_PPWAIT)
strlcat(state, "V", sizeof(state));
if (p->p_flag & P_SYSTEM || p->p_lock > 0)
strlcat(state, "L", sizeof(state));
Avoid NULL pointer dereference, for a process which is not (yet) a member of a process group, e.g. during the system bootstrap. Submitted by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> MFC after: 1 week
2016-04-15 09:13:01 +00:00
if (p->p_pgrp != NULL && p->p_session != NULL &&
SESS_LEADER(p))
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
strlcat(state, "s", sizeof(state));
/* Cheated here and didn't compare pgid's. */
if (p->p_flag & P_CONTROLT)
strlcat(state, "+", sizeof(state));
if (cred != NULL && jailed(cred))
strlcat(state, "J", sizeof(state));
db_printf(" %-6.6s ", state);
Put square backets ([]) around process names for system processes to patch the behavior of ps(1).
2009-10-01 19:12:14 +00:00
if (p->p_flag & P_HADTHREADS) {
Use __LP64__ rather than the PTR64 hack. Suggested by: ru
2006-05-11 21:59:55 +00:00
#ifdef __LP64__
Put square backets ([]) around process names for system processes to patch the behavior of ps(1).
2009-10-01 19:12:14 +00:00
db_printf(" (threaded) ");
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#else
Put square backets ([]) around process names for system processes to patch the behavior of ps(1).
2009-10-01 19:12:14 +00:00
db_printf(" (threaded) ");
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#endif
Put square backets ([]) around process names for system processes to patch the behavior of ps(1).
2009-10-01 19:12:14 +00:00
if (p->p_flag & P_SYSTEM)
db_printf("[");
db_printf("%s", p->p_comm);
if (p->p_flag & P_SYSTEM)
db_printf("]");
db_printf("\n");
}
Completely redo thread states. Reviewed by: davidxu@freebsd.org
2002-09-11 08:13:56 +00:00
FOREACH_THREAD_IN_PROC(p, td) {
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
dumpthread(p, td, p->p_flag & P_HADTHREADS);
Simplify the pager support in DDB. Allowing different db commands to install custom pager functions didn't actually happen in practice (they all just used the simple pager and passed in a local quit pointer). So, just hardcode the simple pager as the only pager and make it set a global db_pager_quit flag that db commands can check when the user hits 'q' (or a suitable variant) at the pager prompt. Also, now that it's easy to do so, enable paging by default for all ddb commands. Any command that wishes to honor the quit flag can do so by checking db_pager_quit. Note that the pager can also be effectively disabled by setting $lines to 0. Other fixes: - 'show idt' on i386 and pc98 now actually checks the quit flag and terminates early. - 'show intr' now actually checks the quit flag and terminates early.
2006-07-12 21:22:44 +00:00
if (db_pager_quit)
Update the 'ps', 'show pci', and 'show ktr' ddb commands to use the new pager callout instead of homerolling their own paging facility.
2003-07-31 17:29:42 +00:00
break;
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
}
Commented out locking that would be used in the ps command if locks were used in ddb.
2002-04-11 21:01:34 +00:00
Use macro API for <sys/queue.h>
2000-12-30 22:06:19 +00:00
p = LIST_NEXT(p, p_list);
Make db_ps tolerant of uninitialized struct proc fields. Improved 'more' function to work like you'd expect. Other misc cleanup.
1995-08-20 05:25:11 +00:00
if (p == NULL && np > 0)
Use macro API for <sys/queue.h>
2000-12-30 22:06:19 +00:00
p = LIST_FIRST(&zombproc);
ddb: space/tab fixes. No functional change. MFC after: 3 days
2014-10-11 20:25:19 +00:00
}
ddb ps function from 1.1.5. Moved into its own file and cleaned up a bit by me.
1994-08-27 16:28:45 +00:00
}
Whitespace nits.
2003-06-06 17:19:27 +00:00
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
static void
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
dumpthread(volatile struct proc *p, volatile struct thread *td, int all)
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
{
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
char state[9], wprefix;
const char *wmesg;
void *wchan;
if (all) {
Some cosmetic tweaks: - Right justify 'pid' label. - Move the uid column to the right 2 columns so that the 3 process id columns (pid, ppid, pgrp) are grouped together. - Expand the uid column to 5 chars. - Don't indent the tid for multithreaded processes. Requested by: bde (1, 2, 4)
2006-08-01 15:29:46 +00:00
db_printf("%6d ", td->td_tid);
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
switch (td->td_state) {
case TDS_RUNNING:
snprintf(state, sizeof(state), "Run");
break;
case TDS_RUNQ:
snprintf(state, sizeof(state), "RunQ");
break;
case TDS_CAN_RUN:
snprintf(state, sizeof(state), "CanRun");
break;
case TDS_INACTIVE:
snprintf(state, sizeof(state), "Inactv");
break;
case TDS_INHIBITED:
state[0] = '\0';
if (TD_ON_LOCK(td))
strlcat(state, "L", sizeof(state));
if (TD_IS_SLEEPING(td)) {
if (td->td_flags & TDF_SINTR)
strlcat(state, "S", sizeof(state));
else
strlcat(state, "D", sizeof(state));
}
if (TD_IS_SWAPPED(td))
strlcat(state, "W", sizeof(state));
if (TD_AWAITING_INTR(td))
strlcat(state, "I", sizeof(state));
if (TD_IS_SUSPENDED(td))
strlcat(state, "s", sizeof(state));
if (state[0] != '\0')
break;
default:
snprintf(state, sizeof(state), "???");
}
db_printf(" %-6.6s ", state);
}
wprefix = ' ';
if (TD_ON_LOCK(td)) {
wprefix = '*';
wmesg = td->td_lockname;
wchan = td->td_blocked;
} else if (TD_ON_SLEEPQ(td)) {
wmesg = td->td_wmesg;
wchan = td->td_wchan;
} else if (TD_IS_RUNNING(td)) {
snprintf(state, sizeof(state), "CPU %d", td->td_oncpu);
wmesg = state;
wchan = NULL;
} else {
wmesg = "";
wchan = NULL;
}
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
db_printf("%c%-7.7s ", wprefix, wmesg);
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
if (wchan == NULL)
Use __LP64__ rather than the PTR64 hack. Suggested by: ru
2006-05-11 21:59:55 +00:00
#ifdef __LP64__
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
db_printf("%18s ", "");
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#else
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
db_printf("%10s ", "");
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
#endif
else
In ddb ps, left justify the non-numeric fields 'state' and 'wmesg' and the fixed-width numeric field 'wchan', as in ps(1). They were sort of centered, although the template shows 'state' as right-justified. The `wmesg' field very rarely has a prefix of '*' (for lock names) that is still to the left of the header, and the width of this field is reduced from 8 to 7 (more than 6 is an error). The 'wmesg' and 'wchan' fields are still misnamed and poorly handled. They are named sort of backwards relative to ps(1): - wmesg in ddb = mwchan in ps - wmesg in ddb = wchan in ps (if it is a wait channel name, not a lock name) - wchan in ddb = nwchan in ps ddb ps wastes lots of space for the unimportant 'wchan' field (20 columns altogether on 64-bit arches). ps(1) documents using a compressed format, but the compression only omits leading nybbles of 0 so it has neveqr worked on arches that put the kernel in the top half of the address space. It just avoids wasting space for an 0x prefix.
2016-08-14 15:26:40 +00:00
db_printf("%p ", wchan);
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
if (p->p_flag & P_SYSTEM)
db_printf("[");
if (td->td_name[0] != '\0')
db_printf("%s", td->td_name);
else
db_printf("%s", td->td_proc->p_comm);
if (p->p_flag & P_SYSTEM)
db_printf("]");
db_printf("\n");
}
Whitespace nit.
2003-07-30 20:59:36 +00:00
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
DB_SHOW_COMMAND(thread, db_show_thread)
{
struct thread *td;
Print out the container lock when showing the thread state in DDB. Tested by: benjsc
2008-06-18 20:42:01 +00:00
struct lock_object *lock;
ddb: finish converting boolean values. The replacement started at r283088 was necessarily incomplete without replacing boolean_t with bool. This also involved cleaning some type mismatches and ansifying old C function declarations. Pointed out by: bde Discussed with: bde, ian, jhb
2015-05-21 15:16:18 +00:00
bool comma;
Add td_swinvoltick to track last involuntary context switch Expose in DDB via "show thread." Reviewed by: markj Sponsored by: EMC / Isilon Storage Division
2016-03-25 19:35:29 +00:00
int delta;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
/* Determine which thread to examine. */
if (have_addr)
ddb: stop boolean screaming. TRUE --> true FALSE--> false Hinted by: NetBSD
2015-05-18 22:27:46 +00:00
td = db_lookup_thread(addr, false);
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
else
td = kdb_thread;
Print out the container lock when showing the thread state in DDB. Tested by: benjsc
2008-06-18 20:42:01 +00:00
lock = (struct lock_object *)td->td_lock;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf("Thread %d at %p:\n", td->td_tid, td);
Make KSE a kernel option, turned on by default in all GENERIC kernel configs except sun4v (which doesn't process signals properly with KSE). Reviewed by: davidxu@
2006-10-26 21:42:22 +00:00
db_printf(" proc (pid %d): %p\n", td->td_proc->p_pid, td->td_proc);
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
if (td->td_name[0] != '\0')
db_printf(" name: %s\n", td->td_name);
Print the stack bounds of the thread.
2007-10-16 17:52:59 +00:00
db_printf(" stack: %p-%p\n", (void *)td->td_kstack,
(void *)(td->td_kstack + td->td_kstack_pages * PAGE_SIZE - 1));
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf(" flags: %#x ", td->td_flags);
db_printf(" pflags: %#x\n", td->td_pflags);
db_printf(" state: ");
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
switch (td->td_state) {
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
case TDS_INACTIVE:
db_printf("INACTIVE\n");
break;
case TDS_CAN_RUN:
db_printf("CAN RUN\n");
break;
case TDS_RUNQ:
db_printf("RUNQ\n");
break;
case TDS_RUNNING:
db_printf("RUNNING (CPU %d)\n", td->td_oncpu);
break;
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
case TDS_INHIBITED:
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf("INHIBITED: {");
ddb: stop boolean screaming. TRUE --> true FALSE--> false Hinted by: NetBSD
2015-05-18 22:27:46 +00:00
comma = false;
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
if (TD_IS_SLEEPING(td)) {
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf("SLEEPING");
ddb: stop boolean screaming. TRUE --> true FALSE--> false Hinted by: NetBSD
2015-05-18 22:27:46 +00:00
comma = true;
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
}
if (TD_IS_SUSPENDED(td)) {
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
if (comma)
db_printf(", ");
db_printf("SUSPENDED");
ddb: stop boolean screaming. TRUE --> true FALSE--> false Hinted by: NetBSD
2015-05-18 22:27:46 +00:00
comma = true;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
}
if (TD_IS_SWAPPED(td)) {
if (comma)
db_printf(", ");
db_printf("SWAPPED");
ddb: stop boolean screaming. TRUE --> true FALSE--> false Hinted by: NetBSD
2015-05-18 22:27:46 +00:00
comma = true;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
}
if (TD_ON_LOCK(td)) {
if (comma)
db_printf(", ");
db_printf("LOCK");
ddb: stop boolean screaming. TRUE --> true FALSE--> false Hinted by: NetBSD
2015-05-18 22:27:46 +00:00
comma = true;
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
}
if (TD_AWAITING_INTR(td)) {
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
if (comma)
db_printf(", ");
db_printf("IWAIT");
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
}
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf("}\n");
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
break;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
default:
db_printf("??? (%#x)\n", td->td_state);
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
break;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
}
if (TD_ON_LOCK(td))
db_printf(" lock: %s turnstile: %p\n", td->td_lockname,
td->td_blocked);
if (TD_ON_SLEEPQ(td))
Rewrite subr_sleepqueue.c use of callouts to not depend on the specifics of callout KPI. Esp., do not depend on the exact interface of callout_stop(9) return values. The main change is that instead of requiring precise callouts, code maintains absolute time to wake up. Callouts now should ensure that a wake occurs at the requested moment, but we can tolerate both run-away callout, and callout_stop(9) lying about running callout either way. As consequence, it removes the constant source of the bugs where sleepq_check_timeout() causes uninterruptible thread state where the thread is detached from CPU, see e.g. r234952 and r296320. Patch also removes dual meaning of the TDF_TIMEOUT flag, making code (IMO much) simpler to reason about. Tested by: pho Reviewed by: jhb Sponsored by: The FreeBSD Foundation MFC after: 1 month Differential revision: https://reviews.freebsd.org/D7137
2016-07-28 09:09:55 +00:00
db_printf(
" wmesg: %s wchan: %p sleeptimo %lx. %jx (curr %lx. %jx)\n",
td->td_wmesg, td->td_wchan,
(long)sbttobt(td->td_sleeptimo).sec,
(uintmax_t)sbttobt(td->td_sleeptimo).frac,
(long)sbttobt(sbinuptime()).sec,
(uintmax_t)sbttobt(sbinuptime()).frac);
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf(" priority: %d\n", td->td_priority);
Print out the container lock when showing the thread state in DDB. Tested by: benjsc
2008-06-18 20:42:01 +00:00
db_printf(" container lock: %s (%p)\n", lock->lo_name, lock);
Add td_swinvoltick to track last involuntary context switch Expose in DDB via "show thread." Reviewed by: markj Sponsored by: EMC / Isilon Storage Division
2016-03-25 19:35:29 +00:00
if (td->td_swvoltick != 0) {
delta = (u_int)ticks - (u_int)td->td_swvoltick;
db_printf(" last voluntary switch: %d ms ago\n",
1000 * delta / hz);
}
if (td->td_swinvoltick != 0) {
delta = (u_int)ticks - (u_int)td->td_swinvoltick;
db_printf(" last involuntary switch: %d ms ago\n",
1000 * delta / hz);
}
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
}
DB_SHOW_COMMAND(proc, db_show_proc)
{
struct thread *td;
struct proc *p;
Simplify the pager support in DDB. Allowing different db commands to install custom pager functions didn't actually happen in practice (they all just used the simple pager and passed in a local quit pointer). So, just hardcode the simple pager as the only pager and make it set a global db_pager_quit flag that db commands can check when the user hits 'q' (or a suitable variant) at the pager prompt. Also, now that it's easy to do so, enable paging by default for all ddb commands. Any command that wishes to honor the quit flag can do so by checking db_pager_quit. Note that the pager can also be effectively disabled by setting $lines to 0. Other fixes: - 'show idt' on i386 and pc98 now actually checks the quit flag and terminates early. - 'show intr' now actually checks the quit flag and terminates early.
2006-07-12 21:22:44 +00:00
int i;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
/* Determine which process to examine. */
if (have_addr)
p = db_lookup_proc(addr);
else
p = kdb_thread->td_proc;
db_printf("Process %d (%s) at %p:\n", p->p_pid, p->p_comm, p);
db_printf(" state: ");
switch (p->p_state) {
case PRS_NEW:
db_printf("NEW\n");
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
break;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
case PRS_NORMAL:
db_printf("NORMAL\n");
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
break;
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
case PRS_ZOMBIE:
db_printf("ZOMBIE\n");
Handle the TDS_INACTIVE state by printing '[INACTIVE]' instead of panic'ing. Also, for unknown thread states, print out the value rather than panic. Panic'ing in the debugger is pointless at best.
2003-06-06 17:20:47 +00:00
break;
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
default:
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf("??? (%#x)\n", p->p_state);
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
}
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
if (p->p_ucred != NULL) {
db_printf(" uid: %d gids: ", p->p_ucred->cr_uid);
for (i = 0; i < p->p_ucred->cr_ngroups; i++) {
db_printf("%d", p->p_ucred->cr_groups[i]);
if (i < (p->p_ucred->cr_ngroups - 1))
db_printf(", ");
}
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
db_printf("\n");
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
}
if (p->p_pptr != NULL)
db_printf(" parent: pid %d at %p\n", p->p_pptr->p_pid,
p->p_pptr);
if (p->p_leader != NULL && p->p_leader != p)
db_printf(" leader: pid %d at %p\n", p->p_leader->p_pid,
p->p_leader);
if (p->p_sysent != NULL)
db_printf(" ABI: %s\n", p->p_sysent->sv_name);
Fix ddb "show proc" to show full arguments PR: 200052 Submitted by: Chang-Hsien Tsai <luke.tw AT gmail.com>
2016-08-01 22:41:50 +00:00
if (p->p_args != NULL) {
db_printf(" arguments: ");
for (i = 0; i < (int)p->p_args->ar_length; i++) {
if (p->p_args->ar_args[i] == '\0')
db_printf(" ");
else
db_printf("%c", p->p_args->ar_args[i]);
}
db_printf("\n");
}
Extend show proc with reaper, sigparent, and vmspace information I have regularly needed the last couple of months. Sponsored by: iXsystems, Inc.
2018-05-25 13:59:48 +00:00
db_printf(" repear: %p reapsubtree: %d\n",
p->p_reaper, p->p_reapsubtree);
db_printf(" sigparent: %d\n", p->p_sigparent);
db_printf(" vmspace: %p\n", p->p_vmspace);
db_printf(" (map %p)\n",
(p->p_vmspace != NULL) ? &p->p_vmspace->vm_map : 0);
db_printf(" (map.pmap %p)\n",
(p->p_vmspace != NULL) ? &p->p_vmspace->vm_map.pmap : 0);
db_printf(" (pmap %p)\n",
(p->p_vmspace != NULL) ? &p->p_vmspace->vm_pmap : 0);
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
db_printf(" threads: %d\n", p->p_numthreads);
FOREACH_THREAD_IN_PROC(p, td) {
dumpthread(p, td, 1);
Simplify the pager support in DDB. Allowing different db commands to install custom pager functions didn't actually happen in practice (they all just used the simple pager and passed in a local quit pointer). So, just hardcode the simple pager as the only pager and make it set a global db_pager_quit flag that db commands can check when the user hits 'q' (or a suitable variant) at the pager prompt. Also, now that it's easy to do so, enable paging by default for all ddb commands. Any command that wishes to honor the quit flag can do so by checking db_pager_quit. Note that the pager can also be effectively disabled by setting $lines to 0. Other fixes: - 'show idt' on i386 and pc98 now actually checks the quit flag and terminates early. - 'show intr' now actually checks the quit flag and terminates early.
2006-07-12 21:22:44 +00:00
if (db_pager_quit)
- Overhaul the 'ps' command in ddb to be mostly readable again. :) It is now back to using fixed-size columns for output and each line of output should fit in 80 columns on both 32-bit and 64-bit architectures. In general the output is close to that of the userland ps(1) with the exception that the 'wmesg' field is mostly similar to the "state" field in top(1) in that it will show either a wmesg, a lock name (prefixed with an *), "CPU xx" (for a running thread), or nothing if none of those three conditions are true. It also respects td_name when listing threads in a multithreaded process. There is a somewhat evilly-defined PTR64 macro I use to make account for the change in the size of the 'wchan' column in the formatted output (wchan is now the only pointer in the ps output and is available so it can be passed to 'show sleepq', 'show turnstile', or 'show lock'). - Add two new commands "show proc [process]" and "show thread [thread]" that show details about the specified process or thread (specified either by pid/tid or pointer), respectively. If an address it not specified, it uses the current kdb thread.
2006-04-25 20:34:04 +00:00
break;
}
Add code to ddb to allow backtracing an arbitrary thread. (show thread {address}) Remove the IDLE kse state and replace it with a change in the way threads sahre KSEs. Every KSE now has a thread, which is considered its "owner" however a KSE may also be lent to other threads in the same group to allow completion of in-kernel work. n this case the owner remains the same and the KSE will revert to the owner when the other work has been completed. All creations of upcalls etc. is now done from kse_reassign() which in turn is called from mi_switch or thread_exit(). This means that special code can be removed from msleep() and cv_wait(). kse_release() does not leave a KSE with no thread any more but converts the existing thread into teh KSE's owner, and sets it up for doing an upcall. It is just inhibitted from being scheduled until there is some reason to do an upcall. Remove all trace of the kse_idle queue since it is no-longer needed. "Idle" KSEs are now on the loanable queue.
2002-12-28 01:23:07 +00:00
}
Add 'findstack' ddb command to search either the thread kernel stack or cached stack containing the specified kernel virtual address. Discussed with: pho MFC after: 1 week
2011-12-16 11:44:20 +00:00
void
ddb: finish converting boolean values. The replacement started at r283088 was necessarily incomplete without replacing boolean_t with bool. This also involved cleaning some type mismatches and ansifying old C function declarations. Pointed out by: bde Discussed with: bde, ian, jhb
2015-05-21 15:16:18 +00:00
db_findstack_cmd(db_expr_t addr, bool have_addr, db_expr_t dummy3 __unused,
char *dummy4 __unused)
Add 'findstack' ddb command to search either the thread kernel stack or cached stack containing the specified kernel virtual address. Discussed with: pho MFC after: 1 week
2011-12-16 11:44:20 +00:00
{
struct proc *p;
struct thread *td;
struct kstack_cache_entry *ks_ce;
vm_offset_t saddr;
if (have_addr)
saddr = addr;
else {
db_printf("Usage: findstack <address>\n");
return;
}
Use FOREACH_PROC_IN_SYSTEM instead of using its unrolled form. Reviewed by: kib
2011-12-18 15:36:21 +00:00
FOREACH_PROC_IN_SYSTEM(p) {
Add 'findstack' ddb command to search either the thread kernel stack or cached stack containing the specified kernel virtual address. Discussed with: pho MFC after: 1 week
2011-12-16 11:44:20 +00:00
FOREACH_THREAD_IN_PROC(p, td) {
if (td->td_kstack <= saddr && saddr < td->td_kstack +
PAGE_SIZE * td->td_kstack_pages) {
db_printf("Thread %p\n", td);
return;
}
}
}
for (ks_ce = kstack_cache; ks_ce != NULL;
ks_ce = ks_ce->next_ks_entry) {
if ((vm_offset_t)ks_ce <= saddr && saddr < (vm_offset_t)ks_ce +
Make kstack_pages a tunable on arm, x86, and powepc. On i386, the initial thread stack is not adjusted by the tunable, the stack is allocated too early to get access to the kernel environment. See TD0_KSTACK_PAGES for the thread0 stack sizing on i386. The tunable was tested on x86 only. From the visual inspection, it seems that it might work on arm and powerpc. The arm USPACE_SVC_STACK_TOP and powerpc USPACE macros seems to be already incorrect for the threads with non-default kstack size. I only changed the macros to use variable instead of constant, since I cannot test. On arm64, mips and sparc64, some static data structures are sized by KSTACK_PAGES, so the tunable is disabled. Sponsored by: The FreeBSD Foundation MFC after: 2 week
2015-08-10 17:18:21 +00:00
PAGE_SIZE * kstack_pages) {
Add 'findstack' ddb command to search either the thread kernel stack or cached stack containing the specified kernel virtual address. Discussed with: pho MFC after: 1 week
2011-12-16 11:44:20 +00:00
db_printf("Cached stack %p\n", ks_ce);
return;
}
}
}
Reference in New Issue Copy Permalink