[PATCH] wait.2: Add ESRCH for when pid == INT_MIN

[PATCH] wait.2: Add ESRCH for when pid == INT_MIN

[Richard Palethorpe]

Please see upstream commit:

 commit dd83c161fbcc5d8be637ab159c0de015cbff5ba4
 Author: zhongjiang <zhongjiang@huawei.com>
 Date:   Mon Jul 10 15:53:01 2017 -0700

     kernel/exit.c: avoid undefined behaviour when calling wait4()

It avoids negating INT_MIN by returning early with ESRCH.

Signed-off-by: Richard Palethorpe <rpalethorpe@suse.com>
---
 man2/wait.2 | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/man2/wait.2 b/man2/wait.2
index 5ae955b3f..63d872ee6 100644
--- a/man2/wait.2
+++ b/man2/wait.2
@@ -455,6 +455,15 @@ was caught; see
 The
 .I options
 argument was invalid.
+.TP
+.B ESRCH
+(for
+.BR wait ()
+or
+.BR waitpid ())
+.I pid
+is equal to
+.I INT_MIN
 .SH CONFORMING TO
 SVr4, 4.3BSD, POSIX.1-2001.
 .SH NOTES
-- 
2.31.1

Re: [PATCH] wait.2: Add ESRCH for when pid == INT_MIN

[Alejandro Colomar (man-pages)]

Hello Richard,

On 7/8/21 12:08 PM, Richard Palethorpe wrote:
> Please see upstream commit:
> 
>  commit dd83c161fbcc5d8be637ab159c0de015cbff5ba4
>  Author: zhongjiang <zhongjiang@huawei.com>
>  Date:   Mon Jul 10 15:53:01 2017 -0700
> 
>      kernel/exit.c: avoid undefined behaviour when calling wait4()
> 
> It avoids negating INT_MIN by returning early with ESRCH.
> 
> Signed-off-by: Richard Palethorpe <rpalethorpe@suse.com>

Patch applied.

Thanks,

Alex

> ---
>  man2/wait.2 | 9 +++++++++
>  1 file changed, 9 insertions(+)
> 
> diff --git a/man2/wait.2 b/man2/wait.2
> index 5ae955b3f..63d872ee6 100644
> --- a/man2/wait.2
> +++ b/man2/wait.2
> @@ -455,6 +455,15 @@ was caught; see
>  The
>  .I options
>  argument was invalid.
> +.TP
> +.B ESRCH
> +(for
> +.BR wait ()
> +or
> +.BR waitpid ())
> +.I pid
> +is equal to
> +.I INT_MIN
>  .SH CONFORMING TO
>  SVr4, 4.3BSD, POSIX.1-2001.
>  .SH NOTES
> 

-- 
Alejandro Colomar
Linux man-pages comaintainer; https://www.kernel.org/doc/man-pages/
http://www.alejandro-colomar.es/

Re: [PATCH] wait.2: Add ESRCH for when pid == INT_MIN

[Jakub Wilk]

* Richard Palethorpe <rpalethorpe@suse.com>, 2021-07-08, 11:08:
>Please see upstream commit:
>
> commit dd83c161fbcc5d8be637ab159c0de015cbff5ba4
> Author: zhongjiang <zhongjiang@huawei.com>
> Date:   Mon Jul 10 15:53:01 2017 -0700
>
>     kernel/exit.c: avoid undefined behaviour when calling wait4()
>
>It avoids negating INT_MIN by returning early with ESRCH.

That sounds like a bug in the kernel, though?

POSIX says the error should be ECHILD if "the process group specified by 
pid does not exist".

-- 
Jakub Wilk

Re: [PATCH] wait.2: Add ESRCH for when pid == INT_MIN

[Richard Palethorpe]

Hello Jakub,

Jakub Wilk <jwilk@jwilk.net> writes:

> * Richard Palethorpe <rpalethorpe@suse.com>, 2021-07-08, 11:08:
>>Please see upstream commit:
>>
>> commit dd83c161fbcc5d8be637ab159c0de015cbff5ba4
>> Author: zhongjiang <zhongjiang@huawei.com>
>> Date:   Mon Jul 10 15:53:01 2017 -0700
>>
>>     kernel/exit.c: avoid undefined behaviour when calling wait4()
>>
>>It avoids negating INT_MIN by returning early with ESRCH.
>
> That sounds like a bug in the kernel, though?
>
> POSIX says the error should be ECHILD if "the process group specified
> by pid does not exist".

The absolute value of INT_MIN is undefined or "not representable" in
two's complement. So I think this can reasonably be considered undefined
behaviour and the kernel can do what it wants.

Also, as the error code is different, we can detect if the fix has been
applied without UBSAN enabled (unlike a similar fix in kill).

OTOH, I would have probably used ECHILD for consistency. However it is
done now and has been in use for some years.

-- 
Thank you,
Richard.