In the current thread group implementation getpid() returns the
thread group leader. There is currently no way for a task in a
thread group to find its true pid.
This patch provides a new system call gettid() (get task id), which
returns the true pid of the task. This is needed in some multi-threaded
apps and libraries.
Dave McCracken
======================================================================
Dave McCracken IBM Linux Base Kernel Team 1-512-838-3059
[email protected] T/L 678-3059
==========
--- linux-2.4.10/arch/i386/kernel/entry.S Sat Sep 8 14:02:32 2001
+++ linux-2.4.10-gettid/arch/i386/kernel/entry.S Mon Oct 8 09:57:39 2001
@@ -619,6 +619,7 @@
.long SYMBOL_NAME(sys_madvise)
.long SYMBOL_NAME(sys_getdents64) /* 220 */
.long SYMBOL_NAME(sys_fcntl64)
+ .long SYMBOL_NAME(sys_gettid)
.long SYMBOL_NAME(sys_ni_syscall) /* reserved for TUX */
.rept NR_syscalls-(.-sys_call_table)/4
--- linux-2.4.10/arch/alpha/kernel/entry.S Wed Jul 25 19:11:05 2001
+++ linux-2.4.10-gettid/arch/alpha/kernel/entry.S Mon Oct 8 09:58:43 2001
@@ -10,7 +10,7 @@
#define SIGCHLD 20
-#define NR_SYSCALLS 378
+#define NR_SYSCALLS 379
/*
* These offsets must match with alpha_mv in <asm/machvec.h>.
@@ -1145,3 +1145,4 @@
.quad sys_mincore /* 375 */
.quad sys_pciconfig_iobase
.quad sys_getdents64
+ .quad sys_gettid
--- linux-2.4.10/arch/sparc/kernel/systbls.S Sun Aug 13 14:01:54 2000
+++ linux-2.4.10-gettid/arch/sparc/kernel/systbls.S Mon Oct 8 09:58:43 2001
@@ -165,5 +165,6 @@
.long sunos_nosys, sunos_nosys
/*250*/ .long sunos_nosys, sunos_nosys, sunos_nosys
.long sunos_nosys, sunos_nosys, sunos_nosys
+ .long sys_gettid
#endif
--- linux-2.4.10/arch/mips/kernel/syscalls.h Mon Jul 2 15:56:40 2001
+++ linux-2.4.10-gettid/arch/mips/kernel/syscalls.h Mon Oct 8 09:58:43 2001
@@ -235,3 +235,4 @@
SYS(sys_madvise, 3)
SYS(sys_getdents64, 3)
SYS(sys_fcntl64, 3) /* 4220 */
+SYS(sys_gettid, 0)
--- linux-2.4.10/arch/ppc/kernel/misc.S Tue Aug 28 08:58:33 2001
+++ linux-2.4.10-gettid/arch/ppc/kernel/misc.S Mon Oct 8 09:58:43 2001
@@ -1120,6 +1120,7 @@
.long sys_fcntl64 /* 204 */
.long sys_madvise /* 205 */
.long sys_mincore /* 206 */
+ .long sys_gettid /* 207 */
.rept NR_syscalls-(.-sys_call_table)/4
.long sys_ni_syscall
.endr
--- linux-2.4.10/arch/m68k/kernel/entry.S Tue Feb 13 16:13:43 2001
+++ linux-2.4.10-gettid/arch/m68k/kernel/entry.S Mon Oct 8 09:58:43 2001
@@ -646,6 +646,7 @@
.long SYMBOL_NAME(sys_ni_syscall)
.long SYMBOL_NAME(sys_ni_syscall)
.long SYMBOL_NAME(sys_getdents64) /* 220 */
+ .long SYMBOL_NAME(sys_gettid)
.rept NR_syscalls-(.-SYMBOL_NAME(sys_call_table))/4
.long SYMBOL_NAME(sys_ni_syscall)
--- linux-2.4.10/arch/sparc64/kernel/systbls.S Wed Aug 23 11:30:13 2000
+++ linux-2.4.10-gettid/arch/sparc64/kernel/systbls.S Mon Oct 8 09:58:43
2001
@@ -225,5 +225,6 @@
.word sunos_nosys, sunos_nosys
/*250*/ .word sunos_nosys, sunos_nosys, sunos_nosys
.word sunos_nosys, sunos_nosys, sys_aplib
+ .word sys_gettid
#endif
--- linux-2.4.10/arch/arm/kernel/calls.S Wed Jun 27 16:12:04 2001
+++ linux-2.4.10-gettid/arch/arm/kernel/calls.S Mon Oct 8 09:58:43 2001
@@ -236,6 +236,7 @@
.long SYMBOL_NAME(sys_mincore)
/* 220 */ .long SYMBOL_NAME(sys_madvise)
.long SYMBOL_NAME(sys_fcntl64)
+ .long SYMBOL_NAME(sys_gettid)
__syscall_end:
.rept NR_syscalls - (__syscall_end - __syscall_start) / 4
--- linux-2.4.10/arch/sh/kernel/entry.S Sat Sep 8 14:29:09 2001
+++ linux-2.4.10-gettid/arch/sh/kernel/entry.S Mon Oct 8 09:58:43 2001
@@ -1298,6 +1298,7 @@
.long SYMBOL_NAME(sys_madvise)
.long SYMBOL_NAME(sys_getdents64) /* 220 */
.long SYMBOL_NAME(sys_fcntl64)
+ .long SYMBOL_NAME(sys_gettid)
/*
* NOTE!! This doesn't have to be exact - we just have
@@ -1305,7 +1306,7 @@
* entries. Don't panic if you notice that this hasn't
* been shrunk every time we add a new system call.
*/
- .rept NR_syscalls-221
+ .rept NR_syscalls-222
.long SYMBOL_NAME(sys_ni_syscall)
.endr
--- linux-2.4.10/arch/mips64/kernel/scall_64.S Sun Sep 9 12:43:01 2001
+++ linux-2.4.10-gettid/arch/mips64/kernel/scall_64.S Mon Oct 8 09:58:43
2001
@@ -347,3 +347,4 @@
PTR sys_mincore
PTR sys_madvise
PTR sys_getdents64
+ PTR sys_gettid
--- linux-2.4.10/arch/mips64/kernel/scall_o32.S Sun Sep 9 12:43:01 2001
+++ linux-2.4.10-gettid/arch/mips64/kernel/scall_o32.S Mon Oct 8 09:58:43
2001
@@ -454,6 +454,7 @@
sys sys_madvise 3
sys sys_getdents64 3
sys sys32_fcntl64 3 /* 4220 */
+ sys sys_gettid 0
.endm
.macro sys function, nargs
--- linux-2.4.10/arch/s390/kernel/entry.S Sun Aug 12 12:38:47 2001
+++ linux-2.4.10-gettid/arch/s390/kernel/entry.S Mon Oct 8 09:58:43 2001
@@ -559,7 +559,8 @@
.long sys_madvise
.long sys_getdents64 /* 220 */
.long sys_fcntl64
- .rept 255-221
+ .long sys_gettid
+ .rept 255-222
.long sys_ni_syscall
.endr
--- linux-2.4.10/arch/parisc/kernel/syscall.S Tue Dec 5 14:29:39 2000
+++ linux-2.4.10-gettid/arch/parisc/kernel/syscall.S Mon Oct 8 09:58:43
2001
@@ -552,6 +552,7 @@
ENTRY_UHOH(shmctl) /* 195 */
ENTRY_SAME(ni_syscall) /* streams1 */
ENTRY_SAME(ni_syscall) /* streams2 */
+ ENTRY_SAME(gettid)
.end
--- linux-2.4.10/arch/cris/kernel/entry.S Thu Jul 26 17:10:06 2001
+++ linux-2.4.10-gettid/arch/cris/kernel/entry.S Mon Oct 8 09:58:43 2001
@@ -966,6 +966,7 @@
.long SYMBOL_NAME(sys_madvise)
.long SYMBOL_NAME(sys_getdents64) /* 220 */
.long SYMBOL_NAME(sys_fcntl64)
+ .long SYMBOL_NAME(sys_gettid)
.long SYMBOL_NAME(sys_ni_syscall) /* reserved for TUX */
/*
@@ -975,7 +976,7 @@
* been shrunk every time we add a new system call.
*/
- .rept NR_syscalls-222
+ .rept NR_syscalls-223
.long SYMBOL_NAME(sys_ni_syscall)
.endr
--- linux-2.4.10/arch/s390x/kernel/entry.S Sun Aug 12 12:38:48 2001
+++ linux-2.4.10-gettid/arch/s390x/kernel/entry.S Mon Oct 8 09:58:43 2001
@@ -594,7 +594,8 @@
.long SYSCALL(sys_madvise,sys32_madvise_wrapper)
.long SYSCALL(sys_getdents64,sys32_getdents64_wrapper)/* 220 */
.long SYSCALL(sys_ni_syscall,sys32_fcntl64_wrapper)
- .rept 255-221
+ .long SYSCALL(sys_gettid,sys_gettid)
+ .rept 255-222
.long SYSCALL(sys_ni_syscall,sys_ni_syscall)
.endr
--- linux-2.4.10/include/asm-i386/unistd.h Fri Aug 11 16:39:23 2000
+++ linux-2.4.10-gettid/include/asm-i386/unistd.h Mon Oct 8 09:58:43 2001
@@ -227,6 +227,7 @@
#define __NR_madvise1 219 /* delete when C lib stub is removed */
#define __NR_getdents64 220
#define __NR_fcntl64 221
+#define __NR_gettid 222
/* user-visible error numbers are in the range -1 - -124: see
<asm-i386/errno.h> */
--- linux-2.4.10/include/asm-mips/unistd.h Mon Jul 2 15:56:40 2001
+++ linux-2.4.10-gettid/include/asm-mips/unistd.h Mon Oct 8 09:58:43 2001
@@ -233,11 +233,12 @@
#define __NR_madvise (__NR_Linux + 218)
#define __NR_getdents64 (__NR_Linux + 219)
#define __NR_fcntl64 (__NR_Linux + 220)
+#define __NR_gettid (__NR_Linux + 221)
/*
* Offset of the last Linux flavoured syscall
*/
-#define __NR_Linux_syscalls 220
+#define __NR_Linux_syscalls 221
#ifndef _LANGUAGE_ASSEMBLY
--- linux-2.4.10/include/asm-alpha/unistd.h Wed Jan 24 17:16:23 2001
+++ linux-2.4.10-gettid/include/asm-alpha/unistd.h Mon Oct 8 09:58:43 2001
@@ -315,6 +315,7 @@
#define __NR_mincore 375
#define __NR_pciconfig_iobase 376
#define __NR_getdents64 377
+#define __NR_gettid 378
#if defined(__GNUC__)
--- linux-2.4.10/include/asm-m68k/unistd.h Mon Nov 27 19:11:26 2000
+++ linux-2.4.10-gettid/include/asm-m68k/unistd.h Mon Oct 8 09:58:43 2001
@@ -222,6 +222,7 @@
#define __NR_setfsuid32 215
#define __NR_setfsgid32 216
#define __NR_getdents64 220
+#define __NR_gettid 221
/* user-visible error numbers are in the range -1 - -122: see
<asm-m68k/errno.h> */
--- linux-2.4.10/include/asm-sparc/unistd.h Mon Aug 14 15:09:07 2000
+++ linux-2.4.10-gettid/include/asm-sparc/unistd.h Mon Oct 8 09:58:43 2001
@@ -271,6 +271,7 @@
#define __NR_fdatasync 253
#define __NR_nfsservctl 254
#define __NR_aplib 255
+#define __NR_gettid 256
#define _syscall0(type,name) \
type name(void) \
--- linux-2.4.10/include/asm-ppc/unistd.h Mon May 21 17:02:06 2001
+++ linux-2.4.10-gettid/include/asm-ppc/unistd.h Mon Oct 8 09:58:43 2001
@@ -213,6 +213,7 @@
#define __NR_fcntl64 204
#define __NR_madvise 205
#define __NR_mincore 206
+#define __NR_gettid 207
#define __NR(n) #n
--- linux-2.4.10/include/asm-sparc64/unistd.h Mon Aug 14 15:09:08 2000
+++ linux-2.4.10-gettid/include/asm-sparc64/unistd.h Mon Oct 8 09:58:43
2001
@@ -273,6 +273,7 @@
#define __NR_fdatasync 253
#define __NR_nfsservctl 254
#define __NR_aplib 255
+#define __NR_gettid 256
#define _syscall0(type,name) \
type name(void) \
--- linux-2.4.10/include/asm-arm/unistd.h Sun Aug 12 13:14:00 2001
+++ linux-2.4.10-gettid/include/asm-arm/unistd.h Mon Oct 8 09:58:43 2001
@@ -240,6 +240,7 @@
#define __NR_mincore (__NR_SYSCALL_BASE+219)
#define __NR_madvise (__NR_SYSCALL_BASE+220)
#define __NR_fcntl64 (__NR_SYSCALL_BASE+221)
+#define __NR_gettid (__NR_SYSCALL_BASE+222)
/*
* The following SWIs are ARM private.
--- linux-2.4.10/include/asm-sh/unistd.h Mon Oct 2 13:57:34 2000
+++ linux-2.4.10-gettid/include/asm-sh/unistd.h Mon Oct 8 09:58:43 2001
@@ -231,6 +231,7 @@
#define __NR_madvise 219
#define __NR_getdents64 220
#define __NR_fcntl64 221
+#define __NR_gettid 222
/* user-visible error numbers are in the range -1 - -125: see
<asm-sh/errno.h> */
--- linux-2.4.10/include/asm-ia64/unistd.h Tue Jul 31 12:30:09 2001
+++ linux-2.4.10-gettid/include/asm-ia64/unistd.h Mon Oct 8 09:58:43 2001
@@ -205,6 +205,7 @@
#define __NR_clone2 1213
#define __NR_getdents64 1214
#define __NR_getunwind 1215
+#define __NR_gettid 1216
#if !defined(__ASSEMBLY__) && !defined(ASSEMBLER)
--- linux-2.4.10/include/asm-mips64/unistd.h Tue Nov 28 23:42:04 2000
+++ linux-2.4.10-gettid/include/asm-mips64/unistd.h Mon Oct 8 09:58:43 2001
@@ -461,11 +461,12 @@
#define __NR_mincore (__NR_Linux + 211)
#define __NR_madvise (__NR_Linux + 212)
#define __NR_getdents64 (__NR_Linux + 213)
+#define __NR_gettid (__NR_Linux + 214)
/*
* Offset of the last Linux flavoured syscall
*/
-#define __NR_Linux_syscalls 213
+#define __NR_Linux_syscalls 214
#ifndef _LANGUAGE_ASSEMBLY
--- linux-2.4.10/include/asm-s390/unistd.h Tue Feb 13 16:13:44 2001
+++ linux-2.4.10-gettid/include/asm-s390/unistd.h Mon Oct 8 09:58:43 2001
@@ -209,6 +209,7 @@
#define __NR_mincore 218
#define __NR_madvise 219
#define __NR_getdents64 220
+#define __NR_gettid 221
/* user-visible error numbers are in the range -1 - -122: see
<asm-s390/errno.h> */
--- linux-2.4.10/include/asm-parisc/unistd.h Tue Dec 5 14:29:39 2000
+++ linux-2.4.10-gettid/include/asm-parisc/unistd.h Mon Oct 8 09:58:43 2001
@@ -689,8 +689,9 @@
#define __NR_getpmsg (__NR_Linux + 196) /* some people
actually want streams */
#define __NR_putpmsg (__NR_Linux + 197) /* some people
actually want streams */
+#define __NR_gettid (__NR_Linux + 198)
-#define __NR_Linux_syscalls 197
+#define __NR_Linux_syscalls 198
#define HPUX_GATEWAY_ADDR 0xC0000004
#define LINUX_GATEWAY_ADDR 0x100
--- linux-2.4.10/include/asm-cris/unistd.h Tue May 1 18:05:00 2001
+++ linux-2.4.10-gettid/include/asm-cris/unistd.h Mon Oct 8 09:58:43 2001
@@ -227,6 +227,7 @@
#define __NR_madvise 219
#define __NR_getdents64 220
#define __NR_fcntl64 221
+#define __NR_gettid 222
/* XXX - _foo needs to be __foo, while __NR_bar could be _NR_bar. */
#define _syscall0(type,name) \
--- linux-2.4.10/include/asm-s390x/unistd.h Wed Jul 25 16:12:03 2001
+++ linux-2.4.10-gettid/include/asm-s390x/unistd.h Mon Oct 8 09:58:43 2001
@@ -179,6 +179,7 @@
#define __NR_mincore 218
#define __NR_madvise 219
#define __NR_getdents64 220
+#define __NR_gettid 222
/* user-visible error numbers are in the range -1 - -122: see
<asm-s390/errno.h> */
--- linux-2.4.10/kernel/timer.c Tue Jun 12 18:40:11 2001
+++ linux-2.4.10-gettid/kernel/timer.c Mon Oct 8 09:58:43 2001
@@ -794,6 +794,11 @@
#endif
+asmlinkage long sys_gettid(void)
+{
+ return current->pid;
+}
+
asmlinkage long sys_nanosleep(struct timespec *rqtp, struct timespec *rmtp)
{
struct timespec t;
In article <71000000.1002555975@baldur> you wrote:
> --- linux-2.4.10/arch/i386/kernel/entry.S Sat Sep 8 14:02:32 2001
> +++ linux-2.4.10-gettid/arch/i386/kernel/entry.S Mon Oct 8 09:57:39 2001
> @@ -619,6 +619,7 @@
> .long SYMBOL_NAME(sys_madvise)
> .long SYMBOL_NAME(sys_getdents64) /* 220 */
> .long SYMBOL_NAME(sys_fcntl64)
> + .long SYMBOL_NAME(sys_gettid)
> .long SYMBOL_NAME(sys_ni_syscall) /* reserved for TUX */
I think there was a reason why this syscall is marked reserved,
could you add it as Nr 224? (223 is reserved in 2.4.11-pre as well).
Christoph
--
Of course it doesn't work. We've performed a software upgrade.
> This patch provides a new system call gettid() (get task id), which
> returns the true pid of the task. This is needed in some multi-threaded
> apps and libraries.
Really the fix is to back out the stupid getpid hack. The thread group
is known by user space and can be managed by user space
> > .long SYMBOL_NAME(sys_fcntl64)
> > + .long SYMBOL_NAME(sys_gettid)
> > .long SYMBOL_NAME(sys_ni_syscall) /* reserved for TUX */
>
> I think there was a reason why this syscall is marked reserved,
> could you add it as Nr 224? (223 is reserved in 2.4.11-pre as well).
Yep - 223 is provisionally the security() syscall for the LSM project
On Mon, 8 Oct 2001, Alan Cox wrote:
>
> Really the fix is to back out the stupid getpid hack. The thread group
> is known by user space and can be managed by user space
I agree on a purely technical level, but the same could be said about all
the other group ID stuff too, so..
I'll add a gettid(), except I won't be moving reserved system calls
around. I didn't even realize I had removed gettid - it was there in the
2.4.0-test kernels that implemented the signal groups, and I meant to only
revert the unsafe signals stuff..
Linus
> I'll add a gettid(), except I won't be moving reserved system calls
> around. I didn't even realize I had removed gettid - it was there in the
> 2.4.0-test kernels that implemented the signal groups, and I meant to only
> revert the unsafe signals stuff..
Would it make more sense to add a getpid() and make the existing one
gettid() to keep compatibility at its sanest ?
On Mon, 8 Oct 2001, Alan Cox wrote:
>
> Would it make more sense to add a getpid() and make the existing one
> gettid() to keep compatibility at its sanest ?
I don't think compatibility is an issue: programs that are (a) threaded
and (b) use the new thread group interface and (c) care about tid simply
do not exist. For the simple reason that they cannot exist - getpid() was
changed at the same time CLONE_THREAD was added.
So the only compatibility worry would be
- people using non-thread-aware libraries together with a CLONE_THREAD
core thing - which is possible especially if they have thread wrappers.
But if those libraries care about "pid/tid" issues, there's no way that
can have well-defined behaviour anyway ;)
- people who have been playing with CLONE_THREAD, and have apps that
depend on the "pid is the 'classical' pid, not the thread ID"
behaviour. In which case adding a new gettid() is the right thing to
do.
Now, I actually seriously doubt either of those are real issues, and it
probably doesn't matter what we do. But I'd ratehr have a system call
called "getpid()" do what POSIX threads have traditionally done, namely
give the ID of the process group ("tpid" in linux kernel-speak), and have
"gettid()" give the thread ID ("pid" in linux kernel-speak).
Linus
While we're at it, we should probably be exporting the tgid from
/proc/$pid/status.
Index: fs/proc/array.c
===================================================================
RCS file: /inst/cvs/linux/fs/proc/array.c,v
retrieving revision 1.5.2.32
diff -u -r1.5.2.32 array.c
--- fs/proc/array.c 2001/08/07 08:10:59 1.5.2.32
+++ fs/proc/array.c 2001/10/08 21:52:30
@@ -151,12 +151,13 @@
read_lock(&tasklist_lock);
buffer += sprintf(buffer,
"State:\t%s\n"
+ "Tgid:\t%d\n"
"Pid:\t%d\n"
"PPid:\t%d\n"
"TracerPid:\t%d\n"
"Uid:\t%d\t%d\t%d\t%d\n"
"Gid:\t%d\t%d\t%d\t%d\n",
- get_task_state(p),
+ get_task_state(p), p->tgid,
p->pid, p->pid ? p->p_opptr->pid : 0, 0,
p->uid, p->euid, p->suid, p->fsuid,
p->gid, p->egid, p->sgid, p->fsgid);
--
dwmw2
In article <[email protected]>,
Linus Torvalds <[email protected]> wrote:
>Now, I actually seriously doubt either of those are real issues, and it
>probably doesn't matter what we do. But I'd ratehr have a system call
>called "getpid()" do what POSIX threads have traditionally done, namely
>give the ID of the process group ("tpid" in linux kernel-speak), and have
>"gettid()" give the thread ID ("pid" in linux kernel-speak).
The kernel doesn't have pthread support - glibc emulates it using clone().
Shouldn't glibc then not also be the one to do the getpid() / gettid()
switching internally only for applications linked agains libpthread ?
If you want to have thread IDs I think you should go all the way
and make thread groups a real entity like sessions and process groups.
Setsid() would reset the thread group ID like it now also resets
the process group ID. Ditto for setpgrp() I think. A session can
contain several process groups which can contain several thread groups.
There should be a way to signal all processes in a thread group-
perhaps that should even be the default, a signal to any process
in the group should be delivered to all of them. Or perhaps only
for SIGSTOP and SIGKILL, all other signals could be delivered
to the thread group leader.
Anyway, food for thought.
Mike.
--
Move sig.