This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: Fw: Unreviewed patch (m32r-linux support)
- From: Mark Kettenis <kettenis at jive dot nl>
- To: sakamoto dot kei at renesas dot com
- Cc: gdb-patches at sources dot redhat dot com
- Date: Wed, 15 Sep 2004 13:25:25 +0200 (CEST)
- Subject: Re: Fw: Unreviewed patch (m32r-linux support)
- References: <019601c49afa$1b78ce60$5169910a@E5A02646>
From: "Kei Sakamoto" <sakamoto.kei@renesas.com>
Date: Wed, 15 Sep 2004 17:00:46 +0900
Mark,
Would you review the attached patch?
I'm afraid that you are still busy because of moving.
But it seems you are the best person to review it
because I revised it based on your advise.
If you are too busy, please let me know.
Ah yes, unfortunately I've been pretty busy up until now.
Anyway, before I start giving you more ill-guided advise, can I ask
you a few question first?
Are there any plans to support the PTRACE_GETREGS/PTRACE_SETREGS
requests? I really think a modern Linux port should support those
(and only those), but Linux/M32R isn't going to support them, the code
can be simplified even more.
Are there any plans to support a FPU?
Meanwhile I have a few more comments concerning m32r-linux-tdep.c.
m32r_linux_sigtramp_start should use safe_frame_unwind_memory instead
of read_memory_unsigned_integer, otherwise there might be unwanted
interaction between signal handlers and breakpoints. Please take a
look at the recent changes to i386-linux-tdep.c to get an idea how
things should look.
Instead of rolling your own m32r_linux_svr4_fetch_link_map_offsets,
please use the generic svr4_ilp32_fetch_link_map_offsets.
Also you'll have to s/TM_FILE/DEPRECATED_TM_FILE/g in your linux.mt
file, because of some recent changes.
Hopefully, this doesn't discourage you. On the bright side, these
changes will make it easier to maintain the Linux/M32R port in the
future.
Cheers,
Mark