[CRIU] [PATCH 7/8] restore: Add restoration of alternative signal stack

[Andrew Vagin]

On Tue, Jun 18, 2013 at 06:29:52PM +0400, Cyrill Gorcunov wrote:
> On Tue, Jun 18, 2013 at 06:15:58PM +0400, Andrew Vagin wrote:
> > > 
> > > We call sigaltstack in only one case -- on dump, to fetch the value(s)
> > > dumpee has. On restore we have 2 ways
> > > 
> > >  - either call for sigaltstack explicitly in restorer.c right before
> > >    we call for sigreturn
> > You use this way in these patches.
> > > 
> > >  - either pass stack value to sigreturn call and allow the kernel to
> > >    restore stack for us
> > > 
> > > I chose second way and simply tune up sigframe. Sounds reasonable?
> > You are going to implement the second way...
> > 
> > I vote for the second way too.
> 
> Nope :-)

Sorry, it's my fault. Yes, it looks good for me.

> 
> [patch dump: Add dumping of alternative signal stack]
> 
> 	init_thread
> 		...
> 		ret = sys_sigaltstack(NULL, &args->sas);
> 		...
> 
> 	init
> 		...
> 		ret = sys_sigaltstack(NULL, &args->sas);
> 		...
> 
> [patch restore: Add restoration of alternative signal stack]
> 
> 	__export_restore_task
> 		...
> 		if (args->has_sas)
> 			setup_sas(rt_sigframe, &args->sas);
> 
> 	construct_sigframe
> 		...
> 		setup_sas(sigframe, core->sas);
> 
> ie explicit sys_sigaltstack are only on dump. restore simply
> tunes up sigframe but doesn't call for sys_sigaltstack.