This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [PATCH 2/4] catch syscall -- try 4 -- Architecture-dependent part
- From: Daniel Jacobowitz <drow at false dot org>
- To: Sérgio Durigan Júnior <sergiodj at linux dot vnet dot ibm dot com>
- Cc: gdb-patches at sourceware dot org, teawater <teawater at gmail dot com>
- Date: Sun, 1 Feb 2009 14:40:08 -0500
- Subject: Re: [PATCH 2/4] catch syscall -- try 4 -- Architecture-dependent part
- References: <1232929833.26873.23.camel@miki>
You might as well just test for PTRACE_O_TRACESYSGOOD, and enable it,
at the same time as all the other tests. That'll remove a bunch of
new code, including current_ptrace_options. Even if we do need to
delay system call tracing until some point in execution, that does not
apply to PTRACE_O_TRACESYSGOOD; it can be safely set right away since
it only has any effect once you use PTRACE_SYSCALL.
Other than that, everything I saw in this patch I already described in
my review of patch 1 (entry point check, XML syscalls in non-Linux
code).
--
Daniel Jacobowitz
CodeSourcery