comments on locking of task->comm
A message that is not in the archive (sent off-list or not yet received): [email protected]
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 52 lines · Alexander Nyberg
tor 2005-05-05 klockan 17:03 -0700 skrev Linux Kernel Mailing List:
Quoted 51 lines
tree
125d9d7553c5f6dc6ad030e4c829a5bf71ab3ef5
parent291c4a75ce7632ee5c565359fb875ba0597f76be
author Paolo 'Blaisorblade' Giarrusso <[email protected]> Fri, 06 May 2005 06:16:12 -0700 committer Linus Torvalds <[email protected]> Fri, 06 May 2005 06:36:48 -0700[PATCH] comments on locking of
task->commAdd some comments about
task->comm, to explain what it is near its definition and provide some important pointers to its uses.Signed-off-by Paolo 'Blaisorblade' Giarrusso <[email protected]>Signed-off-by Andrew Morton <[email protected]>Signed-off-by Linus Torvalds <[email protected]>exec.c | 4 +++- linux/sched.h | 7 +++++-- 2 files changed, 8 insertions(+), 3 deletions(-)Index:
fs/exec.c
===================================================================
---be0fe48738b481b4b172cc9a98ac799ca79aece2/fs/exec.c(mode:100644 sha1:52acff3f44f09b6e841424deddd6599938eeafdf) +++125d9d7553c5f6dc6ad030e4c829a5bf71ab3ef5/fs/exec.c(mode:100644 sha1:e56ee24370255e2ab4df9a3933ec03f0d07a2de3) @@ -869,9 +869,11 @@ int flush_old_exec(structlinux_binprm*(quoted diff) +3
-1if (current->euid == current->uid && current->egid == current->gid) current->mm->dumpable = 1; name = bprm->filename; + + /* Copies the binary name from after last slash */ for (i=0; (ch = *(name++)) != '\0';) { if (ch == '/') - i = 0; + i = 0; /* overwrite what we wrote */ else if (i < (sizeof(tcomm) - 1)) tcomm[i++] = ch;Index:
include/linux/sched.h
===================================================================
---be0fe48738b481b4b172cc9a98ac799ca79aece2/include/linux/sched.h(mode:100644 sha1:5f868a5885811571fdc701037bf7b09b40a746b8) +++125d9d7553c5f6dc6ad030e4c829a5bf71ab3ef5/include/linux/sched.h(mode:100644 sha1:4dbb109022f3646ff39b7f64464bebb0200071fc) @@ -578,7 +578,7 @@ structtask_struct{(quoted diff) +1
-1unsigned long flags; /* per process flags, defined below */ unsigned long ptrace; - int lock_depth; /* Lock depth */ + int lock_depth; /* BKL lock depth */
Quoted 14 lines
Review struct task_struct { · line 667
Quoted 11 lines
(quoted diff) +4
-1int prio, static_prio; struct list_head run_list;@@ -661,7 +661,10 @@ struct task_struct { struct key *thread_keyring; /* keyring private to this thread */ #endif int oomkilladj; /* OOM kill score adjustment (bit shift). */ - char comm[TASK_COMM_LEN]; + char comm[TASK_COMM_LEN]; /* executable name excluding path + - access with [gs]et_task_comm (which lock + it with task_lock()) + - initialized normally by flush_old_exec */1) comm is not necessarily the executable name, with
sys_prctl=>PR_SET_NAMEyou can change the name arbitrarily
Ok, this can be added in the comment... however that's to give a meaning to it...
Quoted 3 lines
2) cbrowser shows alot of acccesses directly to ->comm, mostly under
current->commbut a few that touch other processes. So this locking comment isn't really valid but may even confuse people.
Well, it's reverse-mapped from the comment about task_lock()... possibly those users don't care or
Quoted 1 line
I really do value comments but I'm not sure about these.
Sorry for not answering promptly, I was busy and overlooked that you mailed me privately (I thought you cc'ed LKML and that nobody really cared, and after I forgot about this).
Signature · 4 lines
Paolo Giarrusso, aka Blaisorblade Skype user "PaoloGiarrusso" Linux registered user n. 292729 http://www.user-mode-linux.org/~blaisorblade