Atsushi Nemoto wrote:
> On Fri, 25 Jul 2008 23:52:33 +0900 (JST), Atsushi Nemoto <anemo@xxxxxxxxxxxxx> wrote:
>>> It seem ok to me to try it. Here is version 3 of the patch, which I was going to send to Ralf.
>> Thanks, it works for me with serial_txx9 kgdboc module.
>
> BTW, is FRAME_POINTER mandatory for kgdb? I agree that FRAME_POINTER
> (ie. -fno-omit-frame-pointer -fno-optimize-sibling-calls) helps source
> level debugging, but I think transparency is more important.
>
> Now kgdboc can be loaded/activated at run-time, so I want to enable
> CONFIG_KGDB usually. But CONFIG_FRAME_POINTER introduces runtime
> overhead on overall kernel, which is too bad (at least on MIPS).
>
> Also, selecting FRAME_POINTER (which is not selectable on MIPS)
> unconditionally looks somewhat inconsistent.
>
> So ... Is this patch reasonable?
>
Sure the patch is reasonable for MIPS, but I think it is worth going a
step further.
There is no technical reason that frame pointers are required for KGDB
in the present mainline sources. This does allow for further
traceability but it is certainly not a requirement for the use of kgdb.
If all you want to do is look at frame 0 and inspect memory or set a
breakpoint and look at some structures kgdb will certainly serve your
purpose.
I'll consider this a defect to the kgdb core and update the
documentation to reflect that it is advised to use frame pointers, but
not a requirement.
Jason.
>From 329f5bb0071eb2a56f2bcf79838ba91784c8b54d Mon Sep 17 00:00:00 2001
From: Jason Wessel <jason.wessel@xxxxxxxxxxxxx>
Date: Mon, 28 Jul 2008 13:48:31 -0500
Subject: [PATCH] kgdb: remove the requirement for CONFIG_FRAME_POINTER
There is no technical reason that the kgdb core requires frame
pointers. It is up to the end user of KGDB to decide if they need
them or not.
[ anemo@xxxxxxxxxxxxx: removed frame pointers on mips ]
Signed-off-by: Jason Wessel <jason.wessel@xxxxxxxxxxxxx>
---
Documentation/DocBook/kgdb.tmpl | 8 ++++++++
lib/Kconfig.kgdb | 11 +++++++----
2 files changed, 15 insertions(+), 4 deletions(-)
diff --git a/Documentation/DocBook/kgdb.tmpl b/Documentation/DocBook/kgdb.tmpl
index e8acd1f..54d3b15 100644
--- a/Documentation/DocBook/kgdb.tmpl
+++ b/Documentation/DocBook/kgdb.tmpl
@@ -98,6 +98,14 @@
"Kernel debugging" select "KGDB: kernel debugging with remote gdb".
</para>
<para>
+ It is advised, but not required that you turn on the
+ CONFIG_FRAME_POINTER kernel option. This option inserts code to
+ into the compiled executable which saves the frame information in
+ registers or on the stack at different points which will allow a
+ debugger such as gdb to more accurately construct stack back traces
+ while debugging the kernel.
+ </para>
+ <para>
Next you should choose one of more I/O drivers to interconnect debugging
host and debugged target. Early boot debugging requires a KGDB
I/O driver that supports early debugging and the driver must be
diff --git a/lib/Kconfig.kgdb b/lib/Kconfig.kgdb
index 2cfd272..9b5d1d7 100644
--- a/lib/Kconfig.kgdb
+++ b/lib/Kconfig.kgdb
@@ -4,14 +4,17 @@ config HAVE_ARCH_KGDB
menuconfig KGDB
bool "KGDB: kernel debugging with remote gdb"
- select FRAME_POINTER
depends on HAVE_ARCH_KGDB
depends on DEBUG_KERNEL && EXPERIMENTAL
help
If you say Y here, it will be possible to remotely debug the
- kernel using gdb. Documentation of kernel debugger is available
- at http://kgdb.sourceforge.net as well as in DocBook form
- in Documentation/DocBook/. If unsure, say N.
+ kernel using gdb. It is recommended but not required, that
+ you also turn on the kernel config option
+ CONFIG_FRAME_POINTER to aid in producing more reliable stack
+ backtraces in the external debugger. Documentation of
+ kernel debugger is available at http://kgdb.sourceforge.net
+ as well as in DocBook form in Documentation/DocBook/. If
+ unsure, say N.
if KGDB
--
1.5.5.1
