|To:||Ralf Baechle <firstname.lastname@example.org>|
|Subject:||Re: [PATCH][MIPS] fix kgdb build error|
|From:||David Daney <email@example.com>|
|Date:||Fri, 24 Oct 2008 10:18:10 -0700|
|Cc:||Yoichi Yuasa <firstname.lastname@example.org>, linux-mips <email@example.com>|
|References:||<firstname.lastname@example.org> <4901F851.email@example.com> <20081024170950.GC25297@linux-mips.org>|
|User-agent:||Thunderbird 126.96.36.199 (X11/20080723)|
Ralf Baechle wrote:
On Fri, Oct 24, 2008 at 09:31:13AM -0700, David Daney wrote:Yoichi Yuasa wrote:ptrace.h needs #include <linux/types.h>[...] Can you try this completely untested patch instead? If it works, I will give it a more thorough test over the next few weeks.This looks correct, I think. Though I was wondering about two special cases: o 32-bit debugger debugging a 64-bit process o 64-bit debugger debugging a 32-bit process The unions make we wonder if that case was considered ...
The ptrace interface only gives access to the raw watch registers, their width and thus the union element used is determined by how the kernel was built (32 or 64 bit).
It is up to the user space debugger to handle address space size issues. The current gdb patch has only been tested on 32 bit systems.
|<Prev in Thread]||Current Thread||[Next in Thread>|
|Previous by Date:||Re: [PATCH][MIPS] fix rb532 build error, Ralf Baechle|
|Next by Date:||[PATCH] rb532: set gpio interrupt status and level for CompactFlash, Florian Fainelli|
|Previous by Thread:||Re: [PATCH][MIPS] fix kgdb build error, Ralf Baechle|
|Next by Thread:||Re: [PATCH][MIPS] fix kgdb build error, Ralf Baechle|
|Indexes:||[Date] [Thread] [Top] [All Lists]|