Re: [PATCH] Incorrect value for SIGRTMAX
From: Andrew Morton (akpm_at_osdl.org)
Date: 01/24/04
- Previous message: Carl-Daniel Hailfinger: "Re: [PATCH] [2.4] forcedeth network driver"
- In reply to: eric.piel_at_tremplin-utc.net: "Re: [PATCH] Incorrect value for SIGRTMAX"
- Next in thread: eric.piel_at_tremplin-utc.net: "Re: [PATCH] Incorrect value for SIGRTMAX"
- Reply: eric.piel_at_tremplin-utc.net: "Re: [PATCH] Incorrect value for SIGRTMAX"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Sat, 24 Jan 2004 14:30:37 -0800 To: eric.piel@tremplin-utc.net
eric.piel@tremplin-utc.net wrote:
>
> Hello,
>
> Few months ago, Corey Minyard corrected handling of incorrect values of signals.
> cf
> http://linux.bkbits.net:8080/linux-2.5/cset@1.1267.56.40?nav=index.html%7Csrc/%7Csrc/kernel%7Crelated/kernel/posix-timers.c
>
> Working on the High-Resolution Timers project, I noticed there is an error in
> good_sigevent() to catch the case when sigev_signo is 0. In this function, we
> want to return NULL when sigev_signo is 0. The one liner attached (used for a
> long time in the HRT patch) should do the trick, it's against vanilla 2.6.1 .
>
OK, the current code is obviously junk:
if ((event->sigev_notify & ~SIGEV_NONE & MIPS_SIGEV) &&
event->sigev_signo &&
((unsigned) (event->sigev_signo > SIGRTMAX)))
return NULL;
a) it's doing
if (foo && foo > N)
which is redundant.
b) it's casting the result of (foo > N) to unsigned which is nonsensical.
Your patch doesn't address b).
I don't thik there's a case in which sigev_signo can be negative anyway.
But if there is, the cast should be done like the below, yes?
kernel/posix-timers.c | 3 +--
1 files changed, 1 insertion(+), 2 deletions(-)
diff -puN kernel/posix-timers.c~SIGRTMAX-fix kernel/posix-timers.c
--- 25/kernel/posix-timers.c~SIGRTMAX-fix 2004-01-24 14:27:13.000000000 -0800
+++ 25-akpm/kernel/posix-timers.c 2004-01-24 14:28:21.000000000 -0800
@@ -344,8 +344,7 @@ static inline struct task_struct * good_
return NULL;
if ((event->sigev_notify & ~SIGEV_NONE & MIPS_SIGEV) &&
- event->sigev_signo &&
- ((unsigned) (event->sigev_signo > SIGRTMAX)))
+ (((unsigned)event->sigev_signo > SIGRTMAX) || !event->sigev_signo))
return NULL;
return rtn;
_
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
- Previous message: Carl-Daniel Hailfinger: "Re: [PATCH] [2.4] forcedeth network driver"
- In reply to: eric.piel_at_tremplin-utc.net: "Re: [PATCH] Incorrect value for SIGRTMAX"
- Next in thread: eric.piel_at_tremplin-utc.net: "Re: [PATCH] Incorrect value for SIGRTMAX"
- Reply: eric.piel_at_tremplin-utc.net: "Re: [PATCH] Incorrect value for SIGRTMAX"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|
