[PATCH 0/4] 'catch syscall' feature -- Introduction

[PATCH 0/4] 'catch syscall' feature -- Introduction

[Sérgio Durigan Júnior]

Hello guys,

This is the new version of the "catch syscall" patch. I have reviewed it
for a while, fixing things that you have proposed to me and implementing
new code. Basically, the changes are:

- Grammar errors fixed (including the "two-spaces-after-dot" rule) (as
pointed out by Michael).

- Rework of the "arch independent" part of the patch, now using the new
breakpoint_ops mechanism (as implemented by Joel).

- Tab-completion support (as suggested by Tromey).

- Storing the syscalls names using XML support, and not directly in the
source-code (as requested by Daniel Jacobowits).

- PPC, PPC64 and i386 support.

- Implementation of the GDB's datadir (Daniel asked this too).

Well, I think that's it. You guys already know what this patch is
supposed to do, but in case you have forgotten:

http://sourceware.org/ml/gdb-patches/2008-09/msg00583.html

Hope you like it :-).

Best regards,

-- 
Sérgio Durigan Júnior
Linux on Power Toolchain - Software Engineer
Linux Technology Center - LTC
IBM Brazil

RE: [PATCH 0/4] 'catch syscall' feature -- Introduction

[Pierre Muller]

  I had just one question about this:

  This patch is linux specific in its current state,
but I thought (but you are welcome to correct me if I am wrong,
as I am unsure of what follows) that other operating systems
also use syscalls (like FreeBSD or NetBSD, probably others also).

  If my previous statement is wrong, you can skip
what is following, otherwise my questions are:
1) Do all OS using syscalls use the same number to
functionality correspondence?
2) Do all OS follow the "linux" habit of never changing the
syscall numbers in changing versions?

If the answer to 2) is NO, it probably means that
extending this patch to such OS's will be harder...

Even if we stick to linux only for now,
shouldn't we explicitly add linux in the name of the processor-dependent
xml files?

This would make it easier to extend the feature to other OS's
later, no?


Pierre Muller
Pascal language support maintainer for GDB



> -----Message d'origine-----
> De : gdb-patches-owner@sourceware.org [mailto:gdb-patches-
> owner@sourceware.org] De la part de Sérgio Durigan Júnior
> Envoyé : Tuesday, November 04, 2008 5:31 AM
> À : gdb-patches@sourceware.org
> Objet : [PATCH 0/4] 'catch syscall' feature -- Introduction
> 
> Hello guys,
> 
> This is the new version of the "catch syscall" patch. I have reviewed
> it
> for a while, fixing things that you have proposed to me and
> implementing
> new code. Basically, the changes are:
> 
> - Grammar errors fixed (including the "two-spaces-after-dot" rule) (as
> pointed out by Michael).
> 
> - Rework of the "arch independent" part of the patch, now using the new
> breakpoint_ops mechanism (as implemented by Joel).
> 
> - Tab-completion support (as suggested by Tromey).
> 
> - Storing the syscalls names using XML support, and not directly in the
> source-code (as requested by Daniel Jacobowits).
> 
> - PPC, PPC64 and i386 support.
> 
> - Implementation of the GDB's datadir (Daniel asked this too).
> 
> Well, I think that's it. You guys already know what this patch is
> supposed to do, but in case you have forgotten:
> 
> http://sourceware.org/ml/gdb-patches/2008-09/msg00583.html
> 
> Hope you like it :-).
> 
> Best regards,
> 
> --
> Sérgio Durigan Júnior
> Linux on Power Toolchain - Software Engineer
> Linux Technology Center - LTC
> IBM Brazil

Re: [PATCH 0/4] 'catch syscall' feature -- Introduction

[Daniel Jacobowitz]

On Tue, Nov 04, 2008 at 09:27:08AM +0100, Pierre Muller wrote:
>   If my previous statement is wrong, you can skip
> what is following, otherwise my questions are:
> 1) Do all OS using syscalls use the same number to
> functionality correspondence?

No.

> 2) Do all OS follow the "linux" habit of never changing the
> syscall numbers in changing versions?

Yes, as far as I know.  This practice is much older than Linux.

-- 
Daniel Jacobowitz
CodeSourcery