Deepak Saxena wrote:
Jason,
Do you have a git tree setup with your latest KGDB work? I have
a pile of patches for ARM and various bug fixes that should be merged
before you try to submit upstream and it'd be good to have a central
repo for coordinating work.
Tnx,
~Deepak
Currently
Sergei,
I looked through the mails that I had from the kgdb-bugreport, and did
not see a formal patch other than to the ethernet drivers. Did you have
a patch for the NET_POLL that you wanted to submit?
I am in the process of updating from 2.6.21rc3 - 2.6.21rc7 and would be
happy to review
Sergei Shtylyov wrote:
That one belonged to [EMAIL PROTECTED] (they're not in a
hurry to apply it though :-).
Well, maybe it's worth to pick them up into KGDB tree -- I'm seeing
there some netpoll changes seem to have never been accepted upstream.
WBR, Sergei
I have submitted the
Tom Rini wrote:
On Thu, May 03, 2007 at 10:45:55AM -0500, Jason Wessel wrote:
- Patch for consideration #1 -
Moved weak definitions to kernel/kgdbarchlib.c to get rid of gcc-4.1 bug.
http://kgdb.cvs.sourceforge.net/kgdb/kgdb-2/core-lite.patch?r1=1.125r2=1.126pathrev=MAIN
Sergei,
This is committed to mips-lite.patch in the linux2_6_21_uprev branch.
The other patches you sent will be reviewed soon.
Thanks,
Jason.
Sergei Shtylyov wrote:
Fix these warnings caused by missing #include in asm-mips/kgdb.h:
In file included from include/linux/kgdb.h:22,
Sergei Shtylyov wrote:
Update KGDB serial driver for Toshiba TX49xx:
- rename it according to the new convention, with _kgdb prefix;
- enable support for the new baud rate option, removing the obsolete #ifdef
cruft;
- enable support for the serial port selection;
- fix write_char
Sergei,
While I would agree with your change, I guess I had not noticed how the
defaults had changed since I originally pushed this upstream.
My original intent was to have the default be config
KGDB_ONLY_MODULES. Then there is no need for the conditionals and
folks who need KGDB as a built
Sergei Shtylyov wrote:
I ask the question of Sergei and anyone else, if it makes more sense
to have the KGDB_ONLY_MODULES as the default?
If yes, I will make the change to have the default as using I/O
modules. If no, it makes sense to merge Sergei's change.
I'd vote against it...
I
Sergei Shtylyov wrote:
In fact, you've brought up an interesting point -- when I tried
using KGDBoE as a module, I failed at that. How did you manage to
make it work?
WBR, Sergei
It used to be the case that you could build a built-in rs232 driver as
well kgdboe and load that as a
Sergei Shtylyov wrote:
Fix errors/warnings in the PXA and AMBA PL011 UARTs' KGDB serial drivers:
- PXA UART selection was missing;
- KGDB baud rate option was not enabled;
- write_char method took argument of type 'int' instead of 'u8'.
Signed-off-by: Vitaly Wool [EMAIL PROTECTED]
Sergei Shtylyov wrote:
[snip snip]
we've been observing different kinds of lockups and even firmware
corruption stepping through functions w/ KGDB when KGDB and
DEBUG_PREEMPT
are enabled in the kernel. It looks like such can't be avoided,
it's just
KGDB and DEBUG_PREEMPT can't safely work
Vitaly Wool wrote:
Hello Jason,
On 5/8/07, Jason Wessel [EMAIL PROTECTED] wrote:
On IA32, I did some experimentation with the latest RT patches to
2.6.21.1 + KGDB modified to use the raw versions of the spin lock and
die_notifier and did not see any particular in a test where I step and
hit
Attached is a patch to fix the 85xx rs232 code to compile correctly.
The plan would be to commit this to the power-lite.patch unless there is
objections.
Jason.
-
This SF.net email is sponsored by DB2 Express
Download
Sergei Shtylyov wrote:
Well, since powerpc-lite.patch removed that callback from
machdep_calls, I've also removed the initializer from
arch/ppc/platforms/85xx/sbc8560.c (and elsewhere). Let me post my
variant?..
By all means :-)
Jason.
Sergei Shtylyov wrote:
Hello.
Jason Wessel wrote:
mips-fix.diff - Updated exception handling for MIPS
mips64_defines.patch - Fix MIPS64 defines for proper compilation
mips64_kgdb_dsubu.patch - Fixed stack pointer corruption in mips64 code
wrs_mips64-usermode-dbg.patch - Usermode debug
Jason Wessel wrote:
Pete/Piet Delaney wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I just did an update from SF with tag linux2_6_21_uprev
Got a few warning and wondered why.
-
-
lib/Kconfig.kgdb
Sergei Shtylyov wrote:
Update KGDB serial driver for CPM UART:
- add limited support for the new baud rate option (by mechanically changing
the related preprocessor directives);
- add the missing UART selection menu, fix the related preprocessor
directives;
- make the driver default
Jason Wessel wrote:
-
Do not infinitely loop hitting a breakpoint breakin
on detach or kill commands issued to KGDB. Instead advance
the PC beyond the breakpoint instruction.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
Committed to arm-lite.patch in linux2_6_21_uprev branch
Jason Wessel wrote:
Fix compile error introduced by powerpc-lite.patch
Signed-off-by: Paul Gortmaker [EMAIL PROTECTED]
---
[snip]
- struct platform_device* pdev = early_uart_get_pdev(co-index);
+ struct platform_device* pdev = early_uart_get_pdev(index);
Committed
Bob Picco wrote:
Jason,
okay. I still want ia64 hardware breakpoint support removed (3rd patch). I
haven't tested it in a couple years. Also it adds code to ivt.S which
needs to be improved.
That is fine by me. I had not seen patch #3, but I'll assume it is on
the way. The first two
Bob Picco wrote:
IA64 doesn't require this interface. So make it conditional for only X86.
Signed-off-by: Bob Picco [EMAIL PROTECTED]
Committed to core-lite.patch in linux2_6_21_uprev.
Jason.
-
This SF.net email is
-by: Jason Wessel [EMAIL PROTECTED]
diff -u linux-2.6.21-rc3/arch/x86_64/kernel/kgdb.c
linux-2.6.21-rc3/arch/x86_64/kernel/kgdb.c
--- linux-2.6.21-rc3/arch/x86_64/kernel/kgdb.c
+++ linux-2.6.21-rc3/arch/x86_64/kernel/kgdb.c
at at -379,7 +379,7 at at
static char intr_desc[] = Stack
Jason Wessel wrote:
mips-fix.diff - Updated exception handling for MIPS
mips64_defines.patch - Fix MIPS64 defines for proper compilation
mips64_kgdb_dsubu.patch - Fixed stack pointer corruption in mips64 code
wrs_mips64-usermode-dbg.patch - Usermode debug exception
forwarding fix
This is committed to powerpc-lite.patch with context changes in
mips-lite.patch and arm-lite.patch in the linux2_6_21_uprev branch.
Jason.
Sergei Shtylyov wrote:
Update KGDB serial driver for MPSC:
- merge the necessary #define's from the serial driver -- the shared header
file
has been
Sergei Shtylyov wrote:
Sanitize the bogus KGDB code for Xilinx ML300:
- remove misplaced calls to early_serial_setup() and kgdb8250_add_port() from
ml300_early_serial_map() -- the wouldn't even compile there;
- #include linux/kgdb.h;
- change CONFIG_KGDB to CONFIG_KGDB_8250;
- enclose
Sergei Shtylyov wrote:
Add KGDB support for Bamboo/Luan PPC4xx platforms:
- #include linux/kgdb.h and add calls the to kgdb8250_add_port();
- remove early_serial_map() method initializers;
- remove all mentions of CONFIG_KGDB.
Signed-off-by: Wade Farnsworth [EMAIL PROTECTED]
Pete/Piet Delaney wrote:
Sounds great to me; avoiding spin locks() is a hassle.
Ever noticed a problem with kgdb surviving a weekend of non-use?
I hit a breakpoint on Saturday and hoped to continue looking at it
today but the kgdb-stub, as usual, got out of phase and I had
to re-do the test
Sergei Shtylyov wrote:
Jason Wessel wrote:
This patch fixes some corner cases where KGDB will silently hang or
kill the system, if a user accidentally tries to source step into a
spin_unlock() call or source step in on a macro containing
smp_processor_id(). The use of raw_smp_processor_id
of sending a trap to unsuspecting code in
the user space (normally resulting in process termination).
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
arch/i386/kernel/kgdb.c | 31 +++
arch/x86_64/kernel/kgdb.c | 17 -
2 files changed, 31
Sergei Shtylyov wrote:
Hello.
Jason Wessel wrote:
If not can you provide me with a test case to see the problem?
Well, for example, with 2.6.18-rt7 kernel, stepping into
smp_processor_id() was blowing away U-Boot (!) on my PPC board (even
in PREEMPT_DESKTOP mode)... On x86, stepping
Jason Wessel wrote:
This patch fixes some corner cases where KGDB will silently hang or
kill the system, if a user accidentally tries to source step into a
spin_unlock() call or source step in on a macro containing
smp_processor_id(). The use of raw_smp_processor_id is desired in
kernel
Jason Wessel wrote:
This patch is to fix another corner case where kgdb can pass a single
step trap to user space which was intended for the kernel. Now a
source level next which single steps over an iret instruction will
cause KGDB to continue and print an error to the console, vs
is request to step from
kernel-user space. If there are no objections, I'll apply this patch
after the review period.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
Index: linux-2.6.21/arch/i386/kernel/kgdb.c
===
--- linux-2.6.21
Sergei Shtylyov wrote:
Well, as it turned out, KGDB was broken WRT h/w breakpoints anyway,
so nobody ever cared it seems. :-)
I meant to say i386-lite.patch. :-)
It does appear to be broken. I merged in the code for playing with the
debug registers and it does not work properly at
Jason Wessel wrote:
Sergei Shtylyov wrote:
Well, as it turned out, KGDB was broken WRT h/w breakpoints anyway,
so nobody ever cared it seems. :-)
I meant to say i386-lite.patch. :-)
It does appear to be broken. I merged in the code for playing with the
debug registers
with
i386_hw_breakpoints.patch and x86_64_breakpoints.patch and merge the
core changes for arch breakpoints to core-lite.patch.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
arch/i386/kernel/kgdb.c| 92 +--
arch/x86_64/kernel/kgdb.c | 117
is to decrement the preempt count by 1 to account for the
exception. With this change user processes no longer die with the
scheduling while atomic error.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
arch/i386/kernel/kgdb.c |6 ++
arch/x86_64/kernel/kgdb.c |6 ++
2 files changed
Sergei Shtylyov wrote:
Add checks for CPM UART being used by KGDB to the console setup() and driver
probe() methods -- previously it was only in the legacy mode path of the
module_init() function. Also, properly indent the related code in that path...
Extracted/modified from the original
Sergei Shtylyov wrote:
Fix some SMP specific issues on MIPS:
- smp_call_function() shouldn't be called with interrupts disabled, but
kgdb_roundup_cpu() fails to actually enable them using local_irq_restore()
as the 'flags' argument passed to it has interrupts already disabled;
-
Sergei Shtylyov wrote:
The serial ports used by kgdb must be initialized very early in the boot.
The ARM 'init_machine' where serial ports are normally initialized is too
late since the 'kgdbwait' command-line option is handled in the early_param()
processing.
Signed-off-by: Kevin Hilman
Jason Wessel wrote:
I rolled up several patches along with my changes to fix HW
breakpoints on x86_64 and i386. I also deprecated some of the
#defines in favor of using the kgdb_ops to control the HW breakpoints
for other non IA archs at the point that these are implemented.
Until
I discovered that the preempt_count problem exists in different forms
due to the way faults are handled are different archs when the
kgdb_may_fault and long jump occur to restore the system context. This
means that a general solution was needed vs a per arch solution.
Attached is the patch
there will only be one for the die handler, but the safest
case is to restore the preempt count to what it was when the context
was saved.
This patch will handle it in an architecture independent way.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
kernel/kgdb.c | 56
Sergei Shtylyov wrote:
The mainline code has the same TASKLET_STATE_SCHED but check and
BUG, yet it didn't seem to give the trace -- I'll investigate today...
Is this during the boot cycle or attaching afterwards?
The former -- it's caused by the 'kgdbwait' option.
Sergei,
I did
in the system,
as there is a race to setup the exception stack vs the schduling
context vs kgdb having the ability to debug the exception.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
kernel/kgdb.c | 15 ---
1 file changed, 8 insertions(+), 7 deletions(-)
Index: linux-2.6.21-standard
Pete/Piet Delaney wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Jason Wessel wrote:
-Original Message-
From: Sergei Shtylyov [mailto:[EMAIL PROTECTED]
+
if(!CHECK_EXCEPTION_STACK()) {
Worth fixing if stmt style while
Sergei Shtylyov wrote:
Hello.
Jason Wessel wrote:
allmodconfig on powerpc (iMac g3) fails due to
git-kgdb.patch. allmodconfig defaults should be changed?
CC arch/powerpc/kernel/kgdb.o
arch/powerpc/kernel/kgdb.c:485:2: error: #error Both XMON and KGDB
selected in .config
Sergei Shtylyov wrote:
Hello.
Wessel, Jason wrote:
Sergei,
kgdb_unset_may_fault() is responsible for restoring the preempt count.
And if it is not in the return path from a set_longjmp, then there is no
way to restore the preempt count.
It may seem counter intuitive to have the mayfault
Tilman Schmidt wrote:
Apart from that, the kernel runs fine, though its subjectively a bit
sluggish compared to 2.6.22-rc6-mm1 without kgdb. Its uname -a is:
Linux xenon 2.6.22-rc6-mm1-test4 #10 SMP PREEMPT Thu Jul 5 23:41:32 CEST 2007
i686 i686 i386 GNU/Linux
I didn't even try to actually
Jan Kiszka wrote:
At this chance... Reminds me that this old issue still seems to be
unsolved in current kgdb:
http://www.mail-archive.com/kgdb-bugreport@lists.sourceforge.net/msg00442.html
I'm only looking at that spot in kgdb right now and /may/ oversee new
border conditions elsewhere.
Bob Picco wrote:
Hi Jasson:
Well ia64 seems to be missing some changes I sent. I must have not done
a quilt refresh before posting my patches to you. This is required to compile
ia64 kgdb.c successfully.
thanks,
bob
Thanks for fixing this Bob. I merged in your changes to the
Rajesham Gajjela wrote:
An interesting bug!!!
with kgdb, on x86_64/linux-2.6.7 or linux-2.6.20,
if some body tries access address such as
0x8000, we get general protection fault,
even though kern_addr_valid() says this is a valid
address.
I just tried to access some invalid
Rajesham Gajjela wrote:
(gdb) p *(char *) (0x80)
Sending packet: $m80,1#8a...Ack
Packet received: E22
Sending packet: $m80,1#8a...Ack
Packet received: E22
Cannot access memory at address 0x80
(gdb)
Try accessing address 0x8000
(not
Rajesham Gajjela wrote:
With the below patch, my bug got fixed. Could you
guys let me know if you have any comments ?
I am of the opinion that its ok to ignore the
events for DIE_GPF.
--- arch/x86_64/kernel/kgdb.c.orig 2007-07-17
07:06:16.463899180 -0700
+++
Sergei Shtylyov wrote:
Since it doesn't make sense both to set kgdb_may_fault flag before setting up
the jump buffer and to resotre the preempt count if a fault hasn't happened,
divorce setting the flag and saving/restoring the count:
- get rid of kgdb_[un]set_may_fault() replacing the calls
Sergei Shtylyov wrote:
Fix the correct_hw_break() method for both i386 and x86_64 to update DR3 and
the corresponding bits in DR7. While at it:
- somewhat clarify the code setting the enable/type/length fileds in DR7;
- initilize the 'correctit' variable right when declaring it;
- fix the
Sergei Shtylyov wrote:
Hello, I wrote:
BTW, I'm having starnge things happening with h/w breakpoints on
i386 -- when I hit initial breakpoint, and enter:
(gdb) hb sys_sync
(gdb) c
and then enter 'sync' after login (even several times), the
breakpoint is not hit. It only starts being
with this implementation vs the prior implementation.
Jason.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
arch/i386/kernel/kgdb.c | 71
--
arch/x86_64/kernel/kgdb.c | 68
include/asm-mips/kdebug.h |1
3
There were several problems with KGDB when no I/O module was attached.
When no I/O module is attached it should correctly exit from the
notifier chain. This patch addresses the problem.
-
This SF.net email is sponsored
I will never understand why source forge eats nearly all attachments...
Jason Wessel wrote:
There were several problems with KGDB when no I/O module was
attached. When no I/O module is attached it should correctly exit
from the notifier chain. This patch addresses the problem
Sergei Shtylyov wrote:
Hello.
Erm... the patches were not separated at this place actually but before
kgdb_arch_handle_exception()... Anyway, x86_64-hw_breakpoints.patch fails to
apply for me now, here's the reject:
Are you sure the source forge cvs was up todate? I don't see
Sergei Shtylyov wrote:
Hm... yet the kernel I was booting was UP. The same happened with
SMP kernel I've just tried. Now I'm confused. :-/
something was broken WRT handling of the h/w breask as being per-CPU
resource but not seeing anything obvious...
The piece that clears DR7 is
PROTECTED]
Signed-off-by: Konstantin Baydarov [EMAIL PROTECTED]
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
include/linux/clocksource.h | 1 +
kernel/kgdb.c | 4
kernel/time/clocksource.c | 12
3 files changed, 17 insertions(+)
Index: linux-2.6.21-standard/include/linux
Konstantin Baydarov wrote:
Hello, Jason. I think you should move definition clocksource_check_watchdog
outside of CONFIG_CLOCKSOURCE_WATCHDOG ifdefs. Because in case when
CONFIG_CLOCKSOURCE_WATCHDOG isn't set kernel with KGDB won't compile.
You are correct. I had committed the old
== pass exception
signal 15 == detach kgdb and pass exception
signal 30 == detach kgdb, set attachwait=0 and pass exception
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
kernel/kgdb.c| 78 ---
lib/Kconfig.kgdb | 10 +++
2 files
returning NOTIFY_DONE to the NMI_IPI if running on the master KGDB
processor.
On x86_64 the nmi_watch dog did not work at all because there was no
die hook for it, so this was added in.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
arch/i386/kernel/kgdb.c|9 -
arch/x86_64/kernel
Jason Wessel wrote:
I have optimized kgdb to not be in the critical path of the softlockup
detection and fixed it to be more robust on the continue. I was
seeing problems from time to time with the latest kernels when using
both the softlockup and the nmi watch dog.
I plan to replace
Pete/Piet Delaney wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Two Questions:
1st: Your wrote:
-
--
This patch implements new behavior to kgdb. Kgdb should have the
ability to wait for a debugger
the
cmpxchg operation.
Jason.
Signed-off-by: Jason Wessel [EMAIL PROTECTED]
---
kernel/kgdb.c | 18 --
1 file changed, 16 insertions(+), 2 deletions(-)
--- a/kernel/kgdb.c
+++ b/kernel/kgdb.c
@@ -121,6 +121,7 @@ struct task_struct *kgdb_usethread, *kgd
int debugger_step
Pete/Piet Delaney wrote:
We are getting a problem with VMware where kernel text is the schedler
is getting wacked with four null bytes into the code. Thought I'd use
the current linux-2.6-kgdb.git tree and possible the CONFIG_DEBUG_RODATA
patch to make kernel text readonly:
Randy,
This patch is fine, and I am committing it to the for_mm kgdb tree.
I am also adding the depends on NET to the KGDBOE_NOMODULE section,
which would otherwise to a select on KGDBOE. We have to cover the case
for KGDB as a module and not as a module.
Thanks,
Jason.
Randy Dunlap wrote:
I have not looked at the fedora kernels, but it would appear that if
there was a shift that some part of the loader or some kernel patch has
done a runtime relocation.
If you knew for absolutely certain that everything was relocated with an
offset, then you could simply load the symbols with an
Derek Atkins wrote:
Hi,
Jason Wessel [EMAIL PROTECTED] writes:
I have not looked at the fedora kernels, but it would appear that if
there was a shift that some part of the loader or some kernel patch has
done a runtime relocation.
Well, it appears to be the case, but I've only
Derek Atkins wrote:
Quoting Jason Wessel [EMAIL PROTECTED]:
Um, okay. How do I do that? My GDB Fu is weak here; how do I
tell gdb that the symbols in vmlinux are all offset? Or how do I
manipulate the vmlinux binary to offset the symbols?
Start gdb with no file. And do something
I merged this patch into the for_mm branch on the kernel org kgdb
repository.
Jason.
[EMAIL PROTECTED] wrote:
From: Randy Dunlap [EMAIL PROTECTED]
Fix kgdb help text typos, grammar, config symbol names, and indentation.
Signed-off-by: Randy Dunlap [EMAIL PROTECTED]
Cc: Jason Wessel [EMAIL
Lance Spaulding wrote:
Randy Dunlap wrote:
The 2.6.23-mm1 patches are here (not as one large patch file, but
as broken-out patch files):
http://www.kernel.org/pub/linux/kernel/people/akpm/patches/2.6/2.6.23/2.6.23-mm1/broken-out/
and the patch-series (tells the order that the patches were
This patch is committed to the 2.6.24 uprev branch.
I also added in the #include for kgdb.h so there are no added warnings.
Thanks,
Jason.
Drew Moseley wrote:
Signed-off-by: Drew Moseley [EMAIL PROTECTED]
Index: linux/arch/arm/mach-omap2/serial.c
I am wondering what branch you are working off of, because there does
not appear to be any instance of CONFIG_MIPS64 in the 2.6.24 uprev
branch which is the latest upstream branch for the kernel.org submission.
It appears to have already been fixed a while back:
The current
Jan Kiszka wrote:
These days, notify_mutex should better be a real mutex.
Signed-off-by: Jan Kiszka [EMAIL PROTECTED]
---
kernel/module.c | 14 +++---
1 file changed, 7 insertions(+), 7 deletions(-)
Index: b/kernel/module.c
Jan Kiszka wrote:
KGDB allows to direct the console output also to the gdb frontend. But
if you switch on CONFIG_KGDB_CONSOLE blindly, you end up without a
suitable initial console for init, causing a boot panic (like I faced:
http://lkml.org/lkml/2008/1/14/284). One workaround is to
Jan Kiszka wrote:
Jason Wessel wrote:
Jan Kiszka wrote:
diff -up arch/x86/kernel/kgdb_32.c arch/x86/kernel/kgdb_64.c
screamed for unification. Here it is.
Signed-off-by: Jan Kiszka [EMAIL PROTECTED]
---
arch/x86/kernel/Makefile_32 |2
arch/x86/kernel/Makefile_64 |2
Jan Kiszka wrote:
As most changes are tightly coupled, this refactoring patch for
KGDB_8250 as well as the core and the new KGDBOC driver comes as a
single chunk. The changes are:
- Reorganized configuration: I/O drivers can be independently
configured as module or built-in
- Dynamic
Jan Kiszka wrote:
Just a beautification of using debugger_active for checking the debugger
state.
Signed-off-by: Jan Kiszka [EMAIL PROTECTED]
---
arch/x86/kernel/kgdb.c |6 +++---
include/linux/kgdb.h |7 ++-
kernel/kgdb.c |8
kernel/sched.c |
Jan Kiszka wrote:
Latest rework of kgdb8250_interrupt caused a regression in that the new
code wrongly assumes a ctrl-c character is also sent on initial connect
from the host. Here is a fix.
Signed-off-by: Jan Kiszka [EMAIL PROTECTED]
---
drivers/serial/8250_kgdb.c |6 +++---
1 file
Jan Kiszka wrote:
Just for the sake of cleanness.
Signed-off-by: Jan Kiszka [EMAIL PROTECTED]
---
lib/Kconfig.kgdb |4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
Index: b/lib/Kconfig.kgdb
===
---
That is probably a good thing to have.
In the latest version of kgdb, the attach wait is completely gone.
Jason.
Sergei Shtylyov wrote:
... but it doesn't for some reason.
---
lib/Kconfig.kgdb |1 +
1 files changed, 1 insertion(+)
Index: linux-2.6/lib/Kconfig.kgdb
I did merge the patch to the 2.6.24 branch, but as mention this code
doesn't exist anymore beyond the 2.6.24 version of kgdb.
Jason.
Sergei Shtylyov wrote:
Hello.
Jason Wessel wrote:
That is probably a good thing to have.
In the latest version of kgdb, the attach wait
Sergei Shtylyov wrote:
Here's the KGDB serial driver for the Alchemy SOC based boards. The baudrate
is
configurable; however, the port number is not, always being designated
DBAU1xx0
boards' debug UART, i.e. UART1 for Au1200 SOC and UART3 on other SOCs. Also,
there is no aynchronous
Himanshu Chauhan wrote:
Dear all,
After making Linux kernel use the DTE port for standard I/O and writing
a KGDB stub driver for the on board DBGU port, I am finally able to
connect to kgdb on my at91rm9200-ek board. When I pass kgdbwait as
parameter, kernel stops after Uncompressing
Amit S. Kale wrote:
On Monday 24 March 2008 07:18:51 pm Jason Wessel wrote:
Amit S. Kale wrote:
On Monday 17 March 2008 11:55:28 pm Konstantin Baydarov wrote:
Problem:
Sometimes(after remote gdb was connected) x86 SMP kernel(with KGDB and
NMI watchdog enabled) hangs when kernel modules
-kgdb.git;a=shortlog;h=for_linus
Thanks,
Jason.
---CLIP HERE---
From: Jason Wessel [EMAIL PROTECTED]
Date: Mon, 21 Apr 2008 15:13:02 -0500
Subject: [PATCH] kgdb: 1000 loops for the single step test
The single step test is not terribly costly and it should be able to
pass at 1000 loops successfully
Jan Engelhardt wrote:
The menuconfig layout was really whacked. Elements that go into a
menuconfig menu must immediately follow the menuconfig object and have a
depends/if on it.
===
commit e302d451d26cce00636b332220b226d6e6f8caee
Author: Jan Engelhardt [EMAIL PROTECTED]
Date: Fri
You can use kgdboe as a kernel built-in or a module regardless of if
your ethernet driver is a module or built-in.
The only restriction is that you cannot using kgdbwait to stop the
kernel prior to starting the user space unless you have kgdboe and the
ethernet driver as builtins. Obviously
Huang Weiyi wrote:
Removed duplicated inlcude linux/delay.h in drivers/misc/kgdbts.c.
Signed-off-by: Huang Weiyi [EMAIL PROTECTED]
--- a/drivers/misc/kgdbts.c 2008-05-10 08:17:30.0 +0800
+++ b/drivers/misc/kgdbts.c 2008-05-10 08:17:54.0 +0800
@@ -102,7 +102,6 @@
Alexander Beregalov wrote:
Hi
I tried to run the latest git kernel and got the following error.
See an attachment for full dmesg.
kgdbts: ERROR PUT: end of test buffer on 'do_fork_test' line 5
expected OK got $E02#a7
[ cut here ]
WARNING: at
/kernel/git/jwessel/kgdb/)
http://kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb/
I was not able to open it.
-Miline
On Sat, Mar 29, 2008 at 12:22 AM, Jason Wessel
[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] wrote:
In the past few months some significant changes and cleanups occurred
It is entirely possible that the kgdb patches for the 2.6.24 kernel do
not work correctly on the ARM architecture.
If you comment out the architecture initialization though, the undefined
instruction handler does not get installed that allows kgdb to function
properly on this architecture.
If
Alexander Beregalov wrote:
2008/5/20 Jason Wessel [EMAIL PROTECTED]:
Alexander Beregalov wrote:
Hi
I tried to run the latest git kernel and got the following error.
See an attachment for full dmesg.
kgdbts: ERROR PUT: end of test buffer on 'do_fork_test' line 5
expected OK got $E02#a7
Thomas Gleixner wrote:
On Wed, 21 May 2008, Jason Wessel wrote:
An RFC type patch follows which addresses the problem, but in my
opinion not in a terribly clean way... Perhaps other folks will
comment on a different or better way to approach the problem or chime
in on if it should be fixed
David Miller wrote:
From: Thomas Gleixner [EMAIL PROTECTED]
Date: Thu, 22 May 2008 23:14:23 +0200 (CEST)
Please use HW breakpoints in the test suite. The RFC patch looks
horrible :)
ftrace needs basically the same kind of facility as kgdb
needs here. So bypassing the need for kgdb won't
1 - 100 of 613 matches
Mail list logo
