2006-10-24 18:37:16

by Oleg Nesterov

Subject: [PATCH] add-process_session-helper-routine-deprecate-old-field-tidy

Cedric Le Goater wrote:
>
> - .session = 1, \
> + .tty_old_pgrp = 0, \
> + { .session = 1}, \

Any reason to initialize .tty_old_pgrp explicitly? This gives a false
positive from grep...

--- rc2-mm2/include/linux/init_task.h~ 2006-10-22 19:28:17.000000000 +0400
+++ rc2-mm2/include/linux/init_task.h 2006-10-24 22:24:04.000000000 +0400
@@ -66,8 +66,7 @@
.cpu_timers = INIT_CPU_TIMERS(sig.cpu_timers), \
.rlim = INIT_RLIMITS, \
.pgrp = 1, \
- .tty_old_pgrp = 0, \
- { .__session = 1}, \
+ { .__session = 1 }, \
}

extern struct nsproxy init_nsproxy;

2006-11-07 16:13:12

by Cédric Le Goater

Subject: Re: [PATCH] add-process_session-helper-routine-deprecate-old-field-tidy

Oleg Nesterov wrote:
> Cedric Le Goater wrote:
>> - .session = 1, \
>> + .tty_old_pgrp = 0, \
>> + { .session = 1}, \
>
> Any reason to initialize .tty_old_pgrp explicitly? This gives a false
> positive from grep...

it helps some version of gcc to find the anonymous union. I need to dig
this issue a little bit more.

thanks,

C.