Re: [PATCH 0/2] MIPS: Move signal return trampolines off the stack.

From: David VomLehn <dvomlehn@cisco.com>
To: David Daney <ddaney@caviumnetworks.com>
Cc: linux-mips@linux-mips.org, Ralf Baechle <ralf@linux-mips.org>
Subject: Re: [PATCH 0/2] MIPS: Move signal return trampolines off the stack.
Date: Wed, 22 Apr 2009 11:31:29 -0700	[thread overview]
Message-ID: <20090422183129.GB9184@cuplxvomd02.corp.sa.net> (raw)
In-Reply-To: <49EF5E58.2040901@caviumnetworks.com>

On Wed, Apr 22, 2009 at 11:13:44AM -0700, David Daney wrote:
> David VomLehn wrote:
...
>> Only one comment, which I would not want to hold up acceptance:
>> based on some numbers sent out recently, it looks like the kernel is
>> experiencing some performance issues with exec() and I think this change will
>> make it slightly slower. You could avoid this by deferring installation of
>> the trampoline to the first use of a system call that registers a signal
>> handler.
>
> I should try to measure this too.  Although this is what x86 et al. do.  
> It is by far much simpler and less prone to bugs that trying to hook  
> into the system calls.  After an executable has had the chance to start  
> additional threads and establish arbitrary mappings things get 
> complicated.

I suspect the overhead is quite small and agree that hooking into the
system calls a bit risky. This might be better done as a phase two, if
at all.

> David Daney

David VomLehn