8:16 a.m.
The first patch allows us to rely on one less heuristic by taking
advantage of information recently made available to us through the
kernel image. The second patch enables analysis of qemu generated
dump files.
v3:
- Forgot the type == 'A' check needed to avoid a ton of unnecessary
strcmps [drew]
v2:
- Drop "arm64: relax symbol filters" patch and just stash the value of
the one absolute symbol we need "_kernel_flags_le". [Dave Anderson]
Also, while working on this, I testing several dumps. The following are
my results.
arm/aarch64 kvm guest kdump testing (P - PASS, F - FAIL). Testing done
with a latest mainline crash utility patched with these patches and a
latest mainline qemu with patches for dump generation.
.-----------------------------------------------------------------------.
| Host | arm32 | arm64 | arm64 | arm64 |
|---------------------------------------|-------|-------|-------|-------|
| Guest | arm32 | arm64 | arm64 | arm32 |
|---------------------------------------|-------|-------|-------|-------|
| Pagesize| 4K | 4K | 64K | 4K |
|=======================================================================|
| kdump in guest | F[1] | P[2] | P[3] | F[1] |
|---------------------------------------|-------|-------|-------|-------|
| qmp-dump-guest-memory <filename>[4] | P | P | P | P |
|---------------------------------------|-------|-------|-------|-------|
| qmp-dump-guest-memory -z <filename>[5]| F[8] | P | P | F[8] |
|---------------------------------------|-------|-------|-------|-------|
| qmp-dump-guest-memory -l <filename>[6]| F[8] | P | P | F[8] |
|---------------------------------------|-------|-------|-------|-------|
| qmp-dump-guest-memory -s <filename>[7]| F[8] | P | P | F[8] |
.-----------------------------------------------------------------------.
[1] Kernel v4.4-rc1 crashes with a NULL pointer dereference at virtual
address 00000000 in a memcpy (crash_kexec/machine_kexec/fncpy/memcpy).
Needs kernel debugging.
[2] Not sure about mainline, but works with the RHEL kernel,
makedumpfile does not yet support arm64 with 4K pages, but using
'core_collector cp' in /etc/kdump.conf allows saving an uncompressed
elf file.
[3] Not sure about mainline, but works with the RHEL kernel,
uses makedumpfile, thus generates a makedumpfile formatted file
using zlib compression.
[4] No format specified, creates an uncompressed elf formatted file.
[5] makedumpfile format, with zlib compression
[6] makedumpfile format, with lzo compression
[7] makedumpfile format, with snappy compression
[8] The crash utility doesn't seem to like arm32 dumps in makedumpfile
format. Looks like the physical page bitmap is all zeros? Needs
qemu and crash debugging.
Andrew Jones (2):
arm64: read pagesize
arm/arm64: read elf notes for qemu generated cores
arm.c | 22 ++++++++++++++++++++++
arm64.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++++++++
defs.h | 2 ++
netdump.c | 21 +++++++++++++++++++++
4 files changed, 97 insertions(+)
--
2.4.3
8:16 a.m.
New subject: [PATCH v3 1/2] arm64: read pagesize
Thanks to kernel commit 9d372c9fab34 "arm64: Add page size to the
kernel image header", we don't need a heuristic to determine the
pagesize on arm64. Keep the heuristic for older kernels, but try
reading the header first.
---
arm64.c | 30 ++++++++++++++++++++++++++++++
defs.h | 1 +
2 files changed, 31 insertions(+)
diff --git a/arm64.c b/arm64.c
index 5f8e2f6852434..9d03d1d54b574 100644
--- a/arm64.c
+++ b/arm64.c
@@ -19,6 +19,7 @@
#include "defs.h"
#include <elf.h>
+#include <endian.h>
#define NOT_IMPLEMENTED(X) error((X), "%s: function not implemented\n",
__func__)
@@ -97,6 +98,32 @@ arm64_init(int when)
break;
case PRE_GDB:
+ if (!machdep->pagesize) {
+ /*
+ * Kerneldoc Documentation/arm64/booting.txt describes
+ * the kernel image header flags field.
+ */
+ value = machdep->machspec->kernel_flags;
+ value = (value >> 1) & 3;
+
+ switch(value)
+ {
+ case 0:
+ break;
+ case 1:
+ machdep->pagesize = 4096;
+ break;
+ case 2:
+ /* TODO: machdep->pagesize = 16384; */
+ error(FATAL, "16K pages not supported.");
+ break;
+ case 3:
+ machdep->pagesize = 65536;
+ break;
+ }
+
+ }
+
if (!machdep->pagesize &&
kernel_symbol_exists("swapper_pg_dir") &&
kernel_symbol_exists("idmap_pg_dir")) {
@@ -305,6 +332,9 @@ arm64_verify_symbol(const char *name, ulong value, char type)
if (!name || !strlen(name))
return FALSE;
+ if ((type == 'A') && STREQ(name, "_kernel_flags_le"))
+ machdep->machspec->kernel_flags = le64toh(value);
+
if (((type == 'A') || (type == 'a')) && (highest_bit_long(value)
!= 63))
return FALSE;
diff --git a/defs.h b/defs.h
index 7004619017e41..3e5dbd99942b2 100644
--- a/defs.h
+++ b/defs.h
@@ -2944,6 +2944,7 @@ struct machine_specific {
ulong crash_kexec_end;
ulong crash_save_cpu_start;
ulong crash_save_cpu_end;
+ ulong kernel_flags;
};
struct arm64_stackframe {
--
2.4.3
1:19 p.m.
New subject: [PATCH v3 1/2] arm64: read pagesize
----- Original Message -----
Thanks to kernel commit 9d372c9fab34 "arm64: Add page size to
the
kernel image header", we don't need a heuristic to determine the
pagesize on arm64. Keep the heuristic for older kernels, but try
reading the header first.
This looks good -- I also added the new kernel_flags field to arm64_dump_machdep_table()
for display by "help -m". Queued for crash-7.1.4:
https://github.com/crash-utility/crash/commit/2d8895c36a3a533b6555a21c866...
Thanks,
Dave
---
arm64.c | 30 ++++++++++++++++++++++++++++++
defs.h | 1 +
2 files changed, 31 insertions(+)
diff --git a/arm64.c b/arm64.c
index 5f8e2f6852434..9d03d1d54b574 100644
--- a/arm64.c
+++ b/arm64.c
@@ -19,6 +19,7 @@
#include "defs.h"
#include <elf.h>
+#include <endian.h>
#define NOT_IMPLEMENTED(X) error((X), "%s: function not implemented\n",
__func__)
@@ -97,6 +98,32 @@ arm64_init(int when)
break;
case PRE_GDB:
+ if (!machdep->pagesize) {
+ /*
+ * Kerneldoc Documentation/arm64/booting.txt describes
+ * the kernel image header flags field.
+ */
+ value = machdep->machspec->kernel_flags;
+ value = (value >> 1) & 3;
+
+ switch(value)
+ {
+ case 0:
+ break;
+ case 1:
+ machdep->pagesize = 4096;
+ break;
+ case 2:
+ /* TODO: machdep->pagesize = 16384; */
+ error(FATAL, "16K pages not supported.");
+ break;
+ case 3:
+ machdep->pagesize = 65536;
+ break;
+ }
+
+ }
+
if (!machdep->pagesize &&
kernel_symbol_exists("swapper_pg_dir") &&
kernel_symbol_exists("idmap_pg_dir")) {
@@ -305,6 +332,9 @@ arm64_verify_symbol(const char *name, ulong value, char
type)
if (!name || !strlen(name))
return FALSE;
+ if ((type == 'A') && STREQ(name, "_kernel_flags_le"))
+ machdep->machspec->kernel_flags = le64toh(value);
+
if (((type == 'A') || (type == 'a')) &&
(highest_bit_long(value) != 63))
return FALSE;
diff --git a/defs.h b/defs.h
index 7004619017e41..3e5dbd99942b2 100644
--- a/defs.h
+++ b/defs.h
@@ -2944,6 +2944,7 @@ struct machine_specific {
ulong crash_kexec_end;
ulong crash_save_cpu_start;
ulong crash_save_cpu_end;
+ ulong kernel_flags;
};
struct arm64_stackframe {
--
2.4.3
8:16 a.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
QEMU can generate both non-makedumpfile (just elf) and makedumpfile
formatted kdumps. In neither case will crash_notes have prstatus, as
crash_kexec doesn't run in the kernel, however the elf notes will
contain the prstatus, and we can dig them out of there.
---
arm.c | 22 ++++++++++++++++++++++
arm64.c | 22 ++++++++++++++++++++++
defs.h | 1 +
netdump.c | 21 +++++++++++++++++++++
4 files changed, 66 insertions(+)
diff --git a/arm.c b/arm.c
index ffc7c06d0f42c..8fc3d48a8ac89 100644
--- a/arm.c
+++ b/arm.c
@@ -597,6 +597,28 @@ arm_get_crash_notes(void)
note = (Elf32_Nhdr *)buf;
p = buf + sizeof(Elf32_Nhdr);
+ /*
+ * dumpfiles created with qemu won't have crash_notes, but there will
+ * be elf notes.
+ */
+ if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE())) {
+ if (DISKDUMP_DUMPFILE())
+ note = diskdump_get_prstatus_percpu(i);
+ else if (KDUMP_DUMPFILE())
+ note = netdump_get_prstatus_percpu(i);
+ if (note) {
+ /*
+ * SIZE(note_buf) accounts for a "final note", which is a
+ * trailing empty elf note header.
+ */
+ long notesz = SIZE(note_buf) - sizeof(Elf32_Nhdr);
+
+ if (sizeof(Elf32_Nhdr) + roundup(note->n_namesz, 4) +
+ note->n_descsz == notesz)
+ BCOPY((char *)note, buf, notesz);
+ }
+ }
+
if (note->n_type != NT_PRSTATUS) {
error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
goto fail;
diff --git a/arm64.c b/arm64.c
index 9d03d1d54b574..b399d7a5c37fe 100644
--- a/arm64.c
+++ b/arm64.c
@@ -1827,6 +1827,28 @@ arm64_get_crash_notes(void)
note = (Elf64_Nhdr *)buf;
p = buf + sizeof(Elf64_Nhdr);
+ /*
+ * dumpfiles created with qemu won't have crash_notes, but there will
+ * be elf notes.
+ */
+ if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE())) {
+ if (DISKDUMP_DUMPFILE())
+ note = diskdump_get_prstatus_percpu(i);
+ else if (KDUMP_DUMPFILE())
+ note = netdump_get_prstatus_percpu(i);
+ if (note) {
+ /*
+ * SIZE(note_buf) accounts for a "final note", which is a
+ * trailing empty elf note header.
+ */
+ long notesz = SIZE(note_buf) - sizeof(Elf64_Nhdr);
+
+ if (sizeof(Elf64_Nhdr) + roundup(note->n_namesz, 4) +
+ note->n_descsz == notesz)
+ BCOPY((char *)note, buf, notesz);
+ }
+ }
+
if (note->n_type != NT_PRSTATUS) {
error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
goto fail;
diff --git a/defs.h b/defs.h
index 3e5dbd99942b2..8a4045ec78c37 100644
--- a/defs.h
+++ b/defs.h
@@ -5720,6 +5720,7 @@ void dump_registers_for_qemu_mem_dump(void);
void kdump_backup_region_init(void);
void display_regs_from_elf_notes(int, FILE *);
void display_ELF_note(int, int, void *, FILE *);
+void *netdump_get_prstatus_percpu(int);
#define PRSTATUS_NOTE (1)
#define QEMU_NOTE (2)
diff --git a/netdump.c b/netdump.c
index bfa818fdef8e1..3ff49f0577c23 100644
--- a/netdump.c
+++ b/netdump.c
@@ -2345,6 +2345,27 @@ dump_Elf64_Nhdr(Elf64_Off offset, int store)
return len;
}
+void *
+netdump_get_prstatus_percpu(int cpu)
+{
+ int online;
+
+ if ((cpu < 0) || (cpu >= nd->num_prstatus_notes))
+ return NULL;
+
+ /*
+ * If no cpu mapping was done, then there must be
+ * a one-to-one relationship between the number
+ * of online cpus and the number of notes.
+ */
+ if ((online = get_cpus_online()) &&
+ (online == kt->cpus) &&
+ (online != nd->num_prstatus_notes))
+ return NULL;
+
+ return nd->nt_prstatus_percpu[cpu];
+}
+
/*
* Send the request to the proper architecture hander.
*/
--
2.4.3
1:31 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
----- Original Message -----
QEMU can generate both non-makedumpfile (just elf) and makedumpfile
formatted kdumps. In neither case will crash_notes have prstatus, as
crash_kexec doesn't run in the kernel, however the elf notes will
contain the prstatus, and we can dig them out of there.
I don't have a lot of ARM and ARM64 dumpfiles, but just doing a
quick sanity test of your patch, I came across this ARM dumpfile,
which I believe may be a QEMU-generated ELF vmcore. I'm not sure,
but it only has 1 NT_PRSTATUS note for the 1 online cpu (of 5 cpus).
But anyway, note that as expected, it cannot find the registers in the
kernel's uninitialized crash_notes -- here without your patch:
$ crash vmcore.pae vmlinux.pae.gz
crash 7.1.4rc15
Copyright (C) 2002-2014 Red Hat, Inc.
Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
Copyright (C) 1999-2006 Hewlett-Packard Co
Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
Copyright (C) 2005, 2011 NEC Corporation
Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
This program is free software, covered by the GNU General Public License,
and you are welcome to change it and/or distribute copies of it under
certain conditions. Enter "help copying" to see the conditions.
This program has absolutely no warranty. Enter "help warranty" for details.
GNU gdb (GDB) 7.6
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=x86_64-unknown-linux-gnu
--target=arm-elf-linux"...
WARNING: invalid note (n_type != NT_PRSTATUS)
WARNING: cannot retrieve registers for active tasks
KERNEL: vmlinux.pae.gz
DUMPFILE: vmcore.pae
CPUS: 5 [OFFLINE: 4]
DATE: Sun Jun 8 18:27:39 2014
UPTIME: 00:03:22
LOAD AVERAGE: 0.16, 0.16, 0.07
TASKS: 51
NODENAME: buildroot
RELEASE: 3.13.5
VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
MACHINE: armv7l (unknown Mhz)
MEMORY: 256 MB
PANIC: "SysRq : Trigger a crash"
PID: 732
COMMAND: "sh"
TASK: 8bcead00 [THREAD_INFO: 8ad32000]
CPU: 0
STATE: TASK_RUNNING (SYSRQ)
crash> bt -a
PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
bt: WARNING: cannot determine starting stack frame for task 8bcead00
PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
bt: WARNING: cannot determine starting stack frame for task 8bc561c0
PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
bt: WARNING: cannot determine starting stack frame for task 8bc56580
PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
bt: WARNING: cannot determine starting stack frame for task 8bc56940
PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
bt: WARNING: cannot determine starting stack frame for task 8bc56d00
crash>
With your patch applied, it generates a SIGSEGV in arm_get_crash_notes():
$ ./crash /usr/dumps/ARM/vmcore.pae /usr/dumps/ARM/vmlinux.pae.gz
crash 7.1.4rc15
Copyright (C) 2002-2014 Red Hat, Inc.
Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
Copyright (C) 1999-2006 Hewlett-Packard Co
Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
Copyright (C) 2005, 2011 NEC Corporation
Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
This program is free software, covered by the GNU General Public License,
and you are welcome to change it and/or distribute copies of it under
certain conditions. Enter "help copying" to see the conditions.
This program has absolutely no warranty. Enter "help warranty" for details.
GNU gdb (GDB) 7.6
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=x86_64-unknown-linux-gnu
--target=arm-elf-linux"...
Segmentation fault (core dumped)
$
I haven't debugged it other than determining that the "note" it looks to
have found the single note OK, but then upon continuation the next time
through the loop, the "note" pointer is valid at line 597, but your
function sets it back to NULL, and therefore it craps out at line 622:
597 note = (Elf32_Nhdr *)buf;
598 p = buf + sizeof(Elf32_Nhdr);
599
600 /*
601 * dumpfiles created with qemu won't have crash_notes, but
there will
602 * be elf notes.
603 */
604 if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() ||
KDUMP_DUMPFILE())) {
605 if (DISKDUMP_DUMPFILE())
606 note = diskdump_get_prstatus_percpu(i);
607 else if (KDUMP_DUMPFILE())
608 note = netdump_get_prstatus_percpu(i);
609 if (note) {
610 /*
611 * SIZE(note_buf) accounts for a "final
note", which is a
612 * trailing empty elf note header.
613 */
614 long notesz = SIZE(note_buf) -
sizeof(Elf32_Nhdr);
615
616 if (sizeof(Elf32_Nhdr) +
roundup(note->n_namesz, 4) +
617 note->n_descsz == notesz)
618 BCOPY((char *)note, buf, notesz);
619 }
620 }
621
622 if (note->n_type != NT_PRSTATUS) {
623 error(WARNING, "invalid note (n_type !=
NT_PRSTATUS)\n");
624 goto fail;
625 }
Not sure how you want to handle that, probably just bail out the same way
if note becomes NULL?
Dave
---
arm.c | 22 ++++++++++++++++++++++
arm64.c | 22 ++++++++++++++++++++++
defs.h | 1 +
netdump.c | 21 +++++++++++++++++++++
4 files changed, 66 insertions(+)
diff --git a/arm.c b/arm.c
index ffc7c06d0f42c..8fc3d48a8ac89 100644
--- a/arm.c
+++ b/arm.c
@@ -597,6 +597,28 @@ arm_get_crash_notes(void)
note = (Elf32_Nhdr *)buf;
p = buf + sizeof(Elf32_Nhdr);
+ /*
+ * dumpfiles created with qemu won't have crash_notes, but there will
+ * be elf notes.
+ */
+ if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE())) {
+ if (DISKDUMP_DUMPFILE())
+ note = diskdump_get_prstatus_percpu(i);
+ else if (KDUMP_DUMPFILE())
+ note = netdump_get_prstatus_percpu(i);
+ if (note) {
+ /*
+ * SIZE(note_buf) accounts for a "final note", which is a
+ * trailing empty elf note header.
+ */
+ long notesz = SIZE(note_buf) - sizeof(Elf32_Nhdr);
+
+ if (sizeof(Elf32_Nhdr) + roundup(note->n_namesz, 4) +
+ note->n_descsz == notesz)
+ BCOPY((char *)note, buf, notesz);
+ }
+ }
+
if (note->n_type != NT_PRSTATUS) {
error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
goto fail;
diff --git a/arm64.c b/arm64.c
index 9d03d1d54b574..b399d7a5c37fe 100644
--- a/arm64.c
+++ b/arm64.c
@@ -1827,6 +1827,28 @@ arm64_get_crash_notes(void)
note = (Elf64_Nhdr *)buf;
p = buf + sizeof(Elf64_Nhdr);
+ /*
+ * dumpfiles created with qemu won't have crash_notes, but there will
+ * be elf notes.
+ */
+ if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE())) {
+ if (DISKDUMP_DUMPFILE())
+ note = diskdump_get_prstatus_percpu(i);
+ else if (KDUMP_DUMPFILE())
+ note = netdump_get_prstatus_percpu(i);
+ if (note) {
+ /*
+ * SIZE(note_buf) accounts for a "final note", which is a
+ * trailing empty elf note header.
+ */
+ long notesz = SIZE(note_buf) - sizeof(Elf64_Nhdr);
+
+ if (sizeof(Elf64_Nhdr) + roundup(note->n_namesz, 4) +
+ note->n_descsz == notesz)
+ BCOPY((char *)note, buf, notesz);
+ }
+ }
+
if (note->n_type != NT_PRSTATUS) {
error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
goto fail;
diff --git a/defs.h b/defs.h
index 3e5dbd99942b2..8a4045ec78c37 100644
--- a/defs.h
+++ b/defs.h
@@ -5720,6 +5720,7 @@ void dump_registers_for_qemu_mem_dump(void);
void kdump_backup_region_init(void);
void display_regs_from_elf_notes(int, FILE *);
void display_ELF_note(int, int, void *, FILE *);
+void *netdump_get_prstatus_percpu(int);
#define PRSTATUS_NOTE (1)
#define QEMU_NOTE (2)
diff --git a/netdump.c b/netdump.c
index bfa818fdef8e1..3ff49f0577c23 100644
--- a/netdump.c
+++ b/netdump.c
@@ -2345,6 +2345,27 @@ dump_Elf64_Nhdr(Elf64_Off offset, int store)
return len;
}
+void *
+netdump_get_prstatus_percpu(int cpu)
+{
+ int online;
+
+ if ((cpu < 0) || (cpu >= nd->num_prstatus_notes))
+ return NULL;
+
+ /*
+ * If no cpu mapping was done, then there must be
+ * a one-to-one relationship between the number
+ * of online cpus and the number of notes.
+ */
+ if ((online = get_cpus_online()) &&
+ (online == kt->cpus) &&
+ (online != nd->num_prstatus_notes))
+ return NULL;
+
+ return nd->nt_prstatus_percpu[cpu];
+}
+
/*
* Send the request to the proper architecture hander.
*/
--
2.4.3
2:18 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
----- Original Message -----
----- Original Message -----
> QEMU can generate both non-makedumpfile (just elf) and makedumpfile
> formatted kdumps. In neither case will crash_notes have prstatus, as
> crash_kexec doesn't run in the kernel, however the elf notes will
> contain the prstatus, and we can dig them out of there.
I don't have a lot of ARM and ARM64 dumpfiles, but just doing a
quick sanity test of your patch, I came across this ARM dumpfile,
which I believe may be a QEMU-generated ELF vmcore. I'm not sure,
but it only has 1 NT_PRSTATUS note for the 1 online cpu (of 5 cpus).
But anyway, note that as expected, it cannot find the registers in the
kernel's uninitialized crash_notes -- here without your patch:
$ crash vmcore.pae vmlinux.pae.gz
crash 7.1.4rc15
Copyright (C) 2002-2014 Red Hat, Inc.
Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
Copyright (C) 1999-2006 Hewlett-Packard Co
Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
Copyright (C) 2005, 2011 NEC Corporation
Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
This program is free software, covered by the GNU General Public License,
and you are welcome to change it and/or distribute copies of it under
certain conditions. Enter "help copying" to see the conditions.
This program has absolutely no warranty. Enter "help warranty" for
details.
GNU gdb (GDB) 7.6
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=x86_64-unknown-linux-gnu
--target=arm-elf-linux"...
WARNING: invalid note (n_type != NT_PRSTATUS)
WARNING: cannot retrieve registers for active tasks
KERNEL: vmlinux.pae.gz
DUMPFILE: vmcore.pae
CPUS: 5 [OFFLINE: 4]
DATE: Sun Jun 8 18:27:39 2014
UPTIME: 00:03:22
LOAD AVERAGE: 0.16, 0.16, 0.07
TASKS: 51
NODENAME: buildroot
RELEASE: 3.13.5
VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
MACHINE: armv7l (unknown Mhz)
MEMORY: 256 MB
PANIC: "SysRq : Trigger a crash"
PID: 732
COMMAND: "sh"
TASK: 8bcead00 [THREAD_INFO: 8ad32000]
CPU: 0
STATE: TASK_RUNNING (SYSRQ)
crash> bt -a
PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
bt: WARNING: cannot determine starting stack frame for task 8bcead00
PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
bt: WARNING: cannot determine starting stack frame for task 8bc561c0
PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
bt: WARNING: cannot determine starting stack frame for task 8bc56580
PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
bt: WARNING: cannot determine starting stack frame for task 8bc56940
PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
bt: WARNING: cannot determine starting stack frame for task 8bc56d00
crash>
With your patch applied, it generates a SIGSEGV in arm_get_crash_notes():
$ ./crash /usr/dumps/ARM/vmcore.pae /usr/dumps/ARM/vmlinux.pae.gz
crash 7.1.4rc15
Copyright (C) 2002-2014 Red Hat, Inc.
Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
Copyright (C) 1999-2006 Hewlett-Packard Co
Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
Copyright (C) 2005, 2011 NEC Corporation
Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
This program is free software, covered by the GNU General Public License,
and you are welcome to change it and/or distribute copies of it under
certain conditions. Enter "help copying" to see the conditions.
This program has absolutely no warranty. Enter "help warranty" for
details.
GNU gdb (GDB) 7.6
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later
<http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=x86_64-unknown-linux-gnu
--target=arm-elf-linux"...
Segmentation fault (core dumped)
$
I haven't debugged it other than determining that the "note" it looks to
have found the single note OK, but then upon continuation the next time
through the loop, the "note" pointer is valid at line 597, but your
function sets it back to NULL, and therefore it craps out at line 622:
597 note = (Elf32_Nhdr *)buf;
598 p = buf + sizeof(Elf32_Nhdr);
599
600 /*
601 * dumpfiles created with qemu won't have
crash_notes, but there will
602 * be elf notes.
603 */
604 if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() ||
KDUMP_DUMPFILE())) {
605 if (DISKDUMP_DUMPFILE())
606 note =
diskdump_get_prstatus_percpu(i);
607 else if (KDUMP_DUMPFILE())
608 note =
netdump_get_prstatus_percpu(i);
609 if (note) {
610 /*
611 * SIZE(note_buf) accounts for a
"final note", which is a
612 * trailing empty elf note header.
613 */
614 long notesz = SIZE(note_buf) -
sizeof(Elf32_Nhdr);
615
616 if (sizeof(Elf32_Nhdr) +
roundup(note->n_namesz, 4) +
617 note->n_descsz == notesz)
618 BCOPY((char *)note, buf,
notesz);
619 }
620 }
621
622 if (note->n_type != NT_PRSTATUS) {
623 error(WARNING, "invalid note (n_type !=
NT_PRSTATUS)\n");
624 goto fail;
625 }
Not sure how you want to handle that, probably just bail out the same way
if note becomes NULL?
If I add this to arm_get_crash_notes(), just after your new function:
if (!note) {
error(WARNING, "cannot find NT_PRSTATUS note for cpu: %d\n", i);
continue;
}
I get this:
$ ./crash /usr/dumps/ARM/vmcore.pae* /usr/dumps/ARM/vmlinux.pae.gz
crash 7.1.4rc15
Copyright (C) 2002-2014 Red Hat, Inc.
Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
Copyright (C) 1999-2006 Hewlett-Packard Co
Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
Copyright (C) 2005, 2011 NEC Corporation
Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
This program is free software, covered by the GNU General Public License,
and you are welcome to change it and/or distribute copies of it under
certain conditions. Enter "help copying" to see the conditions.
This program has absolutely no warranty. Enter "help warranty" for details.
GNU gdb (GDB) 7.6
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=x86_64-unknown-linux-gnu
--target=arm-elf-linux"...
WARNING: cannot find NT_PRSTATUS note for cpu: 1
WARNING: cannot find NT_PRSTATUS note for cpu: 2
WARNING: cannot find NT_PRSTATUS note for cpu: 3
WARNING: cannot find NT_PRSTATUS note for cpu: 4
KERNEL: /usr/dumps/ARM/vmlinux.pae.gz
DUMPFILE: /usr/dumps/ARM/vmcore.pae
CPUS: 5 [OFFLINE: 4]
DATE: Sun Jun 8 18:27:39 2014
UPTIME: 00:03:22
LOAD AVERAGE: 0.16, 0.16, 0.07
TASKS: 51
NODENAME: buildroot
RELEASE: 3.13.5
VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
MACHINE: armv7l (unknown Mhz)
MEMORY: 256 MB
PANIC: "SysRq : Trigger a crash"
PID: 732
COMMAND: "sh"
TASK: 8bcead00 [THREAD_INFO: 8ad32000]
CPU: 0
STATE: TASK_RUNNING (SYSRQ)
crash> bt -a
PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
#0 [<80265064>] (sysrq_handle_crash) from [<80265810>]
#1 [<80265810>] (__handle_sysrq) from [<80265928>]
#2 [<80265928>] (write_sysrq_trigger) from [<80112120>]
#3 [<80112120>] (proc_reg_write) from [<800c9840>]
#4 [<800c9840>] (vfs_write) from [<800c9be4>]
#5 [<800c9be4>] (sys_write) from [<8000e3e0>]
pc : [<76e9cfdc>] lr : [<0000f998>] psr: 600d0010
sp : 7eab862c ip : 00000000 fp : 000a82a4
r10: 00000020 r9 : 000a8294 r8 : 00000001
r7 : 00000004 r6 : 000a9bf0 r5 : 00000001 r4 : 000a7d88
r3 : 00000000 r2 : 00000002 r1 : 000a9bf0 r0 : 00000001
Flags: nZCv IRQs on FIQs on Mode USER_32 ISA ARM
PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
bt: WARNING: cannot determine starting stack frame for task 8bc561c0
PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
bt: WARNING: cannot determine starting stack frame for task 8bc56580
PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
bt: WARNING: cannot determine starting stack frame for task 8bc56940
PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
bt: WARNING: cannot determine starting stack frame for task 8bc56d00
crash>
Note that if I did "goto fail" instead of "continue", I lose the good
cpu 0 backtrace from the NT_PRSTATUS that your patch found, so doing it
this way is the best of both worlds.
Dave
2:54 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
On Fri, Nov 20, 2015 at 03:18:55PM -0500, Dave Anderson wrote:
----- Original Message -----
>
>
> ----- Original Message -----
> > QEMU can generate both non-makedumpfile (just elf) and makedumpfile
> > formatted kdumps. In neither case will crash_notes have prstatus, as
> > crash_kexec doesn't run in the kernel, however the elf notes will
> > contain the prstatus, and we can dig them out of there.
>
> I don't have a lot of ARM and ARM64 dumpfiles, but just doing a
> quick sanity test of your patch, I came across this ARM dumpfile,
> which I believe may be a QEMU-generated ELF vmcore. I'm not sure,
> but it only has 1 NT_PRSTATUS note for the 1 online cpu (of 5 cpus).
If it's more than a day old then it won't be from qemu. I just posted
the patches for that yesterday morning :-)
>
> But anyway, note that as expected, it cannot find the registers in the
> kernel's uninitialized crash_notes -- here without your patch:
So it looks like there may be other dump types (besides qemu generated)
that can result in missing crash_notes. I don't know what those are, other
than just corruption? In any case, I think this dump is still a good test
case for the reason you found below.
>
> $ crash vmcore.pae vmlinux.pae.gz
>
> crash 7.1.4rc15
> Copyright (C) 2002-2014 Red Hat, Inc.
> Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> Copyright (C) 1999-2006 Hewlett-Packard Co
> Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> Copyright (C) 2005, 2011 NEC Corporation
> Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> This program is free software, covered by the GNU General Public License,
> and you are welcome to change it and/or distribute copies of it under
> certain conditions. Enter "help copying" to see the conditions.
> This program has absolutely no warranty. Enter "help warranty" for
> details.
>
> GNU gdb (GDB) 7.6
> Copyright (C) 2013 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later
> <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law. Type "show
copying"
> and "show warranty" for details.
> This GDB was configured as "--host=x86_64-unknown-linux-gnu
> --target=arm-elf-linux"...
>
> WARNING: invalid note (n_type != NT_PRSTATUS)
> WARNING: cannot retrieve registers for active tasks
>
> KERNEL: vmlinux.pae.gz
> DUMPFILE: vmcore.pae
> CPUS: 5 [OFFLINE: 4]
> DATE: Sun Jun 8 18:27:39 2014
> UPTIME: 00:03:22
> LOAD AVERAGE: 0.16, 0.16, 0.07
> TASKS: 51
> NODENAME: buildroot
> RELEASE: 3.13.5
> VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
> MACHINE: armv7l (unknown Mhz)
> MEMORY: 256 MB
> PANIC: "SysRq : Trigger a crash"
> PID: 732
> COMMAND: "sh"
> TASK: 8bcead00 [THREAD_INFO: 8ad32000]
> CPU: 0
> STATE: TASK_RUNNING (SYSRQ)
>
> crash> bt -a
> PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
> bt: WARNING: cannot determine starting stack frame for task 8bcead00
>
> PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
> bt: WARNING: cannot determine starting stack frame for task 8bc561c0
>
> PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
> bt: WARNING: cannot determine starting stack frame for task 8bc56580
>
> PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
> bt: WARNING: cannot determine starting stack frame for task 8bc56940
>
> PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
> bt: WARNING: cannot determine starting stack frame for task 8bc56d00
> crash>
>
>
> With your patch applied, it generates a SIGSEGV in arm_get_crash_notes():
>
> $ ./crash /usr/dumps/ARM/vmcore.pae /usr/dumps/ARM/vmlinux.pae.gz
>
> crash 7.1.4rc15
> Copyright (C) 2002-2014 Red Hat, Inc.
> Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> Copyright (C) 1999-2006 Hewlett-Packard Co
> Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> Copyright (C) 2005, 2011 NEC Corporation
> Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> This program is free software, covered by the GNU General Public License,
> and you are welcome to change it and/or distribute copies of it under
> certain conditions. Enter "help copying" to see the conditions.
> This program has absolutely no warranty. Enter "help warranty" for
> details.
>
> GNU gdb (GDB) 7.6
> Copyright (C) 2013 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later
> <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law. Type "show
copying"
> and "show warranty" for details.
> This GDB was configured as "--host=x86_64-unknown-linux-gnu
> --target=arm-elf-linux"...
>
> Segmentation fault (core dumped)
> $
>
> I haven't debugged it other than determining that the "note" it looks
to
> have found the single note OK, but then upon continuation the next time
> through the loop, the "note" pointer is valid at line 597, but your
> function sets it back to NULL, and therefore it craps out at line 622:
>
> 597 note = (Elf32_Nhdr *)buf;
> 598 p = buf + sizeof(Elf32_Nhdr);
> 599
> 600 /*
> 601 * dumpfiles created with qemu won't have
> crash_notes, but there will
> 602 * be elf notes.
> 603 */
> 604 if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE()
||
> KDUMP_DUMPFILE())) {
> 605 if (DISKDUMP_DUMPFILE())
> 606 note =
> diskdump_get_prstatus_percpu(i);
> 607 else if (KDUMP_DUMPFILE())
> 608 note =
> netdump_get_prstatus_percpu(i);
> 609 if (note) {
> 610 /*
> 611 * SIZE(note_buf) accounts for a
> "final note", which is a
> 612 * trailing empty elf note header.
> 613 */
> 614 long notesz = SIZE(note_buf) -
> sizeof(Elf32_Nhdr);
> 615
> 616 if (sizeof(Elf32_Nhdr) +
> roundup(note->n_namesz, 4) +
> 617 note->n_descsz == notesz)
> 618 BCOPY((char *)note, buf,
> notesz);
> 619 }
> 620 }
> 621
> 622 if (note->n_type != NT_PRSTATUS) {
> 623 error(WARNING, "invalid note (n_type !=
> NT_PRSTATUS)\n");
> 624 goto fail;
> 625 }
>
> Not sure how you want to handle that, probably just bail out the same way
> if note becomes NULL?
If I add this to arm_get_crash_notes(), just after your new function:
if (!note) {
error(WARNING, "cannot find NT_PRSTATUS note for cpu: %d\n", i);
continue;
}
I get this:
$ ./crash /usr/dumps/ARM/vmcore.pae* /usr/dumps/ARM/vmlinux.pae.gz
crash 7.1.4rc15
Copyright (C) 2002-2014 Red Hat, Inc.
Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
Copyright (C) 1999-2006 Hewlett-Packard Co
Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
Copyright (C) 2005, 2011 NEC Corporation
Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
This program is free software, covered by the GNU General Public License,
and you are welcome to change it and/or distribute copies of it under
certain conditions. Enter "help copying" to see the conditions.
This program has absolutely no warranty. Enter "help warranty" for details.
GNU gdb (GDB) 7.6
Copyright (C) 2013 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law. Type "show copying"
and "show warranty" for details.
This GDB was configured as "--host=x86_64-unknown-linux-gnu
--target=arm-elf-linux"...
WARNING: cannot find NT_PRSTATUS note for cpu: 1
WARNING: cannot find NT_PRSTATUS note for cpu: 2
WARNING: cannot find NT_PRSTATUS note for cpu: 3
WARNING: cannot find NT_PRSTATUS note for cpu: 4
KERNEL: /usr/dumps/ARM/vmlinux.pae.gz
DUMPFILE: /usr/dumps/ARM/vmcore.pae
CPUS: 5 [OFFLINE: 4]
DATE: Sun Jun 8 18:27:39 2014
UPTIME: 00:03:22
LOAD AVERAGE: 0.16, 0.16, 0.07
TASKS: 51
NODENAME: buildroot
RELEASE: 3.13.5
VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
MACHINE: armv7l (unknown Mhz)
MEMORY: 256 MB
PANIC: "SysRq : Trigger a crash"
PID: 732
COMMAND: "sh"
TASK: 8bcead00 [THREAD_INFO: 8ad32000]
CPU: 0
STATE: TASK_RUNNING (SYSRQ)
crash> bt -a
PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
#0 [<80265064>] (sysrq_handle_crash) from [<80265810>]
#1 [<80265810>] (__handle_sysrq) from [<80265928>]
#2 [<80265928>] (write_sysrq_trigger) from [<80112120>]
#3 [<80112120>] (proc_reg_write) from [<800c9840>]
#4 [<800c9840>] (vfs_write) from [<800c9be4>]
#5 [<800c9be4>] (sys_write) from [<8000e3e0>]
pc : [<76e9cfdc>] lr : [<0000f998>] psr: 600d0010
sp : 7eab862c ip : 00000000 fp : 000a82a4
r10: 00000020 r9 : 000a8294 r8 : 00000001
r7 : 00000004 r6 : 000a9bf0 r5 : 00000001 r4 : 000a7d88
r3 : 00000000 r2 : 00000002 r1 : 000a9bf0 r0 : 00000001
Flags: nZCv IRQs on FIQs on Mode USER_32 ISA ARM
PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
bt: WARNING: cannot determine starting stack frame for task 8bc561c0
PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
bt: WARNING: cannot determine starting stack frame for task 8bc56580
PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
bt: WARNING: cannot determine starting stack frame for task 8bc56940
PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
bt: WARNING: cannot determine starting stack frame for task 8bc56d00
crash>
Note that if I did "goto fail" instead of "continue", I lose the good
cpu 0 backtrace from the NT_PRSTATUS that your patch found, so doing it
this way is the best of both worlds.
I agree that the 'if (!note) continue' that you added is a good idea to
try and salvage this type of dump. It shouldn't happen with qemu generated
dumps, but anything's possible when a kernel panics...
Would you like me to spin a v4 with this condition added? Or, since it
actually seems to be addressing a non-qemu-generated dump issue, then
maybe you just want to submit it as a new patch on top of the qemu
patch?
Thanks,
drew
3:25 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
----- Original Message -----
On Fri, Nov 20, 2015 at 03:18:55PM -0500, Dave Anderson wrote:
>
>
> ----- Original Message -----
> >
> >
> > ----- Original Message -----
> > > QEMU can generate both non-makedumpfile (just elf) and makedumpfile
> > > formatted kdumps. In neither case will crash_notes have prstatus, as
> > > crash_kexec doesn't run in the kernel, however the elf notes will
> > > contain the prstatus, and we can dig them out of there.
> >
> > I don't have a lot of ARM and ARM64 dumpfiles, but just doing a
> > quick sanity test of your patch, I came across this ARM dumpfile,
> > which I believe may be a QEMU-generated ELF vmcore. I'm not sure,
> > but it only has 1 NT_PRSTATUS note for the 1 online cpu (of 5 cpus).
If it's more than a day old then it won't be from qemu. I just posted
the patches for that yesterday morning :-)
Even for 32-bit ARM? I honestly can't remember where/who it came from,
although it's in my notes somewhere. Because the non-panicking cpus
are offline, it's possible that it's a kdump-generated dumpfile, and
the other cpu's note were not captured. I don't know, but regardless,
it turns out to be a damn good test case!
> >
> > But anyway, note that as expected, it cannot find the registers in the
> > kernel's uninitialized crash_notes -- here without your patch:
So it looks like there may be other dump types (besides qemu generated)
that can result in missing crash_notes. I don't know what those are, other
than just corruption? In any case, I think this dump is still a good test
case for the reason you found below.
> >
> > $ crash vmcore.pae vmlinux.pae.gz
> >
> > crash 7.1.4rc15
> > Copyright (C) 2002-2014 Red Hat, Inc.
> > Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> > Copyright (C) 1999-2006 Hewlett-Packard Co
> > Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> > Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> > Copyright (C) 2005, 2011 NEC Corporation
> > Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> > Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> > This program is free software, covered by the GNU General Public
> > License,
> > and you are welcome to change it and/or distribute copies of it under
> > certain conditions. Enter "help copying" to see the conditions.
> > This program has absolutely no warranty. Enter "help warranty"
for
> > details.
> >
> > GNU gdb (GDB) 7.6
> > Copyright (C) 2013 Free Software Foundation, Inc.
> > License GPLv3+: GNU GPL version 3 or later
> > <http://gnu.org/licenses/gpl.html>
> > This is free software: you are free to change and redistribute it.
> > There is NO WARRANTY, to the extent permitted by law. Type "show
> > copying"
> > and "show warranty" for details.
> > This GDB was configured as "--host=x86_64-unknown-linux-gnu
> > --target=arm-elf-linux"...
> >
> > WARNING: invalid note (n_type != NT_PRSTATUS)
> > WARNING: cannot retrieve registers for active tasks
> >
> > KERNEL: vmlinux.pae.gz
> > DUMPFILE: vmcore.pae
> > CPUS: 5 [OFFLINE: 4]
> > DATE: Sun Jun 8 18:27:39 2014
> > UPTIME: 00:03:22
> > LOAD AVERAGE: 0.16, 0.16, 0.07
> > TASKS: 51
> > NODENAME: buildroot
> > RELEASE: 3.13.5
> > VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
> > MACHINE: armv7l (unknown Mhz)
> > MEMORY: 256 MB
> > PANIC: "SysRq : Trigger a crash"
> > PID: 732
> > COMMAND: "sh"
> > TASK: 8bcead00 [THREAD_INFO: 8ad32000]
> > CPU: 0
> > STATE: TASK_RUNNING (SYSRQ)
> >
> > crash> bt -a
> > PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
> > bt: WARNING: cannot determine starting stack frame for task 8bcead00
> >
> > PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
> > bt: WARNING: cannot determine starting stack frame for task 8bc561c0
> >
> > PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
> > bt: WARNING: cannot determine starting stack frame for task 8bc56580
> >
> > PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
> > bt: WARNING: cannot determine starting stack frame for task 8bc56940
> >
> > PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
> > bt: WARNING: cannot determine starting stack frame for task 8bc56d00
> > crash>
> >
> >
> > With your patch applied, it generates a SIGSEGV in arm_get_crash_notes():
> >
> > $ ./crash /usr/dumps/ARM/vmcore.pae /usr/dumps/ARM/vmlinux.pae.gz
> >
> > crash 7.1.4rc15
> > Copyright (C) 2002-2014 Red Hat, Inc.
> > Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> > Copyright (C) 1999-2006 Hewlett-Packard Co
> > Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> > Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> > Copyright (C) 2005, 2011 NEC Corporation
> > Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> > Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> > This program is free software, covered by the GNU General Public
> > License,
> > and you are welcome to change it and/or distribute copies of it under
> > certain conditions. Enter "help copying" to see the conditions.
> > This program has absolutely no warranty. Enter "help warranty"
for
> > details.
> >
> > GNU gdb (GDB) 7.6
> > Copyright (C) 2013 Free Software Foundation, Inc.
> > License GPLv3+: GNU GPL version 3 or later
> > <http://gnu.org/licenses/gpl.html>
> > This is free software: you are free to change and redistribute it.
> > There is NO WARRANTY, to the extent permitted by law. Type "show
> > copying"
> > and "show warranty" for details.
> > This GDB was configured as "--host=x86_64-unknown-linux-gnu
> > --target=arm-elf-linux"...
> >
> > Segmentation fault (core dumped)
> > $
> >
> > I haven't debugged it other than determining that the "note" it
looks to
> > have found the single note OK, but then upon continuation the next time
> > through the loop, the "note" pointer is valid at line 597, but your
> > function sets it back to NULL, and therefore it craps out at line 622:
> >
> > 597 note = (Elf32_Nhdr *)buf;
> > 598 p = buf + sizeof(Elf32_Nhdr);
> > 599
> > 600 /*
> > 601 * dumpfiles created with qemu won't have
> > crash_notes, but there will
> > 602 * be elf notes.
> > 603 */
> > 604 if (note->n_namesz == 0 &&
(DISKDUMP_DUMPFILE()
> > ||
> > KDUMP_DUMPFILE())) {
> > 605 if (DISKDUMP_DUMPFILE())
> > 606 note =
> > diskdump_get_prstatus_percpu(i);
> > 607 else if (KDUMP_DUMPFILE())
> > 608 note =
> > netdump_get_prstatus_percpu(i);
> > 609 if (note) {
> > 610 /*
> > 611 * SIZE(note_buf) accounts for a
> > "final note", which is a
> > 612 * trailing empty elf note
> > header.
> > 613 */
> > 614 long notesz = SIZE(note_buf) -
> > sizeof(Elf32_Nhdr);
> > 615
> > 616 if (sizeof(Elf32_Nhdr) +
> > roundup(note->n_namesz, 4) +
> > 617 note->n_descsz == notesz)
> > 618 BCOPY((char *)note, buf,
> > notesz);
> > 619 }
> > 620 }
> > 621
> > 622 if (note->n_type != NT_PRSTATUS) {
> > 623 error(WARNING, "invalid note (n_type !=
> > NT_PRSTATUS)\n");
> > 624 goto fail;
> > 625 }
> >
> > Not sure how you want to handle that, probably just bail out the same way
> > if note becomes NULL?
>
> If I add this to arm_get_crash_notes(), just after your new function:
>
> if (!note) {
> error(WARNING, "cannot find NT_PRSTATUS note for cpu:
%d\n",
> i);
> continue;
> }
>
> I get this:
>
> $ ./crash /usr/dumps/ARM/vmcore.pae* /usr/dumps/ARM/vmlinux.pae.gz
>
> crash 7.1.4rc15
> Copyright (C) 2002-2014 Red Hat, Inc.
> Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> Copyright (C) 1999-2006 Hewlett-Packard Co
> Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> Copyright (C) 2005, 2011 NEC Corporation
> Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> This program is free software, covered by the GNU General Public License,
> and you are welcome to change it and/or distribute copies of it under
> certain conditions. Enter "help copying" to see the conditions.
> This program has absolutely no warranty. Enter "help warranty" for
> details.
>
> GNU gdb (GDB) 7.6
> Copyright (C) 2013 Free Software Foundation, Inc.
> License GPLv3+: GNU GPL version 3 or later
> <http://gnu.org/licenses/gpl.html>
> This is free software: you are free to change and redistribute it.
> There is NO WARRANTY, to the extent permitted by law. Type "show
> copying"
> and "show warranty" for details.
> This GDB was configured as "--host=x86_64-unknown-linux-gnu
> --target=arm-elf-linux"...
>
> WARNING: cannot find NT_PRSTATUS note for cpu: 1
> WARNING: cannot find NT_PRSTATUS note for cpu: 2
> WARNING: cannot find NT_PRSTATUS note for cpu: 3
> WARNING: cannot find NT_PRSTATUS note for cpu: 4
> KERNEL: /usr/dumps/ARM/vmlinux.pae.gz
> DUMPFILE: /usr/dumps/ARM/vmcore.pae
> CPUS: 5 [OFFLINE: 4]
> DATE: Sun Jun 8 18:27:39 2014
> UPTIME: 00:03:22
> LOAD AVERAGE: 0.16, 0.16, 0.07
> TASKS: 51
> NODENAME: buildroot
> RELEASE: 3.13.5
> VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
> MACHINE: armv7l (unknown Mhz)
> MEMORY: 256 MB
> PANIC: "SysRq : Trigger a crash"
> PID: 732
> COMMAND: "sh"
> TASK: 8bcead00 [THREAD_INFO: 8ad32000]
> CPU: 0
> STATE: TASK_RUNNING (SYSRQ)
>
> crash> bt -a
> PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
> #0 [<80265064>] (sysrq_handle_crash) from [<80265810>]
> #1 [<80265810>] (__handle_sysrq) from [<80265928>]
> #2 [<80265928>] (write_sysrq_trigger) from [<80112120>]
> #3 [<80112120>] (proc_reg_write) from [<800c9840>]
> #4 [<800c9840>] (vfs_write) from [<800c9be4>]
> #5 [<800c9be4>] (sys_write) from [<8000e3e0>]
> pc : [<76e9cfdc>] lr : [<0000f998>] psr: 600d0010
> sp : 7eab862c ip : 00000000 fp : 000a82a4
> r10: 00000020 r9 : 000a8294 r8 : 00000001
> r7 : 00000004 r6 : 000a9bf0 r5 : 00000001 r4 : 000a7d88
> r3 : 00000000 r2 : 00000002 r1 : 000a9bf0 r0 : 00000001
> Flags: nZCv IRQs on FIQs on Mode USER_32 ISA ARM
>
> PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
> bt: WARNING: cannot determine starting stack frame for task 8bc561c0
>
> PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
> bt: WARNING: cannot determine starting stack frame for task 8bc56580
>
> PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
> bt: WARNING: cannot determine starting stack frame for task 8bc56940
>
> PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
> bt: WARNING: cannot determine starting stack frame for task 8bc56d00
> crash>
>
> Note that if I did "goto fail" instead of "continue", I lose the
good
> cpu 0 backtrace from the NT_PRSTATUS that your patch found, so doing it
> this way is the best of both worlds.
I agree that the 'if (!note) continue' that you added is a good idea to
try and salvage this type of dump. It shouldn't happen with qemu generated
dumps, but anything's possible when a kernel panics...
Would you like me to spin a v4 with this condition added? Or, since it
actually seems to be addressing a non-qemu-generated dump issue, then
maybe you just want to submit it as a new patch on top of the qemu
patch?
No, I'm cleaning it up now -- there are a couple other gotcha's w/respect
to backtracing, since we're now going to keep the ms->panic_task_regs[cpus]
array intact, even if there's no note in some of them. I'm going to make
sure it's zero-filled in the not-found case, and I have to make sure that
it's recognized on both architectures. I'll post the the updated additions
to arm.c and arm64.c so that you can do a quick test on a legitimate
QEMU dump.
Might be Monday, I'm not sure...
Thanks,
Dave
3:43 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
On Fri, Nov 20, 2015 at 04:25:38PM -0500, Dave Anderson wrote:
----- Original Message -----
> On Fri, Nov 20, 2015 at 03:18:55PM -0500, Dave Anderson wrote:
> >
> >
> > ----- Original Message -----
> > >
> > >
> > > ----- Original Message -----
> > > > QEMU can generate both non-makedumpfile (just elf) and makedumpfile
> > > > formatted kdumps. In neither case will crash_notes have prstatus, as
> > > > crash_kexec doesn't run in the kernel, however the elf notes
will
> > > > contain the prstatus, and we can dig them out of there.
> > >
> > > I don't have a lot of ARM and ARM64 dumpfiles, but just doing a
> > > quick sanity test of your patch, I came across this ARM dumpfile,
> > > which I believe may be a QEMU-generated ELF vmcore. I'm not sure,
> > > but it only has 1 NT_PRSTATUS note for the 1 online cpu (of 5 cpus).
>
> If it's more than a day old then it won't be from qemu. I just posted
> the patches for that yesterday morning :-)
Even for 32-bit ARM? I honestly can't remember where/who it came from,
Yup. The qemu patches I posted are for both arm and aarch64. Before those
patches it wasn't possible to generate a dump with qemu for either.
although it's in my notes somewhere. Because the non-panicking
cpus
are offline, it's possible that it's a kdump-generated dumpfile, and
the other cpu's note were not captured. I don't know, but regardless,
it turns out to be a damn good test case!
>
> > >
> > > But anyway, note that as expected, it cannot find the registers in the
> > > kernel's uninitialized crash_notes -- here without your patch:
>
> So it looks like there may be other dump types (besides qemu generated)
> that can result in missing crash_notes. I don't know what those are, other
> than just corruption? In any case, I think this dump is still a good test
> case for the reason you found below.
>
> > >
> > > $ crash vmcore.pae vmlinux.pae.gz
> > >
> > > crash 7.1.4rc15
> > > Copyright (C) 2002-2014 Red Hat, Inc.
> > > Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> > > Copyright (C) 1999-2006 Hewlett-Packard Co
> > > Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> > > Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> > > Copyright (C) 2005, 2011 NEC Corporation
> > > Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> > > Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> > > This program is free software, covered by the GNU General Public
> > > License,
> > > and you are welcome to change it and/or distribute copies of it under
> > > certain conditions. Enter "help copying" to see the
conditions.
> > > This program has absolutely no warranty. Enter "help
warranty" for
> > > details.
> > >
> > > GNU gdb (GDB) 7.6
> > > Copyright (C) 2013 Free Software Foundation, Inc.
> > > License GPLv3+: GNU GPL version 3 or later
> > > <http://gnu.org/licenses/gpl.html>
> > > This is free software: you are free to change and redistribute it.
> > > There is NO WARRANTY, to the extent permitted by law. Type "show
> > > copying"
> > > and "show warranty" for details.
> > > This GDB was configured as "--host=x86_64-unknown-linux-gnu
> > > --target=arm-elf-linux"...
> > >
> > > WARNING: invalid note (n_type != NT_PRSTATUS)
> > > WARNING: cannot retrieve registers for active tasks
> > >
> > > KERNEL: vmlinux.pae.gz
> > > DUMPFILE: vmcore.pae
> > > CPUS: 5 [OFFLINE: 4]
> > > DATE: Sun Jun 8 18:27:39 2014
> > > UPTIME: 00:03:22
> > > LOAD AVERAGE: 0.16, 0.16, 0.07
> > > TASKS: 51
> > > NODENAME: buildroot
> > > RELEASE: 3.13.5
> > > VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
> > > MACHINE: armv7l (unknown Mhz)
> > > MEMORY: 256 MB
> > > PANIC: "SysRq : Trigger a crash"
> > > PID: 732
> > > COMMAND: "sh"
> > > TASK: 8bcead00 [THREAD_INFO: 8ad32000]
> > > CPU: 0
> > > STATE: TASK_RUNNING (SYSRQ)
> > >
> > > crash> bt -a
> > > PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
> > > bt: WARNING: cannot determine starting stack frame for task 8bcead00
> > >
> > > PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
> > > bt: WARNING: cannot determine starting stack frame for task 8bc561c0
> > >
> > > PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
> > > bt: WARNING: cannot determine starting stack frame for task 8bc56580
> > >
> > > PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
> > > bt: WARNING: cannot determine starting stack frame for task 8bc56940
> > >
> > > PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
> > > bt: WARNING: cannot determine starting stack frame for task 8bc56d00
> > > crash>
> > >
> > >
> > > With your patch applied, it generates a SIGSEGV in arm_get_crash_notes():
> > >
> > > $ ./crash /usr/dumps/ARM/vmcore.pae /usr/dumps/ARM/vmlinux.pae.gz
> > >
> > > crash 7.1.4rc15
> > > Copyright (C) 2002-2014 Red Hat, Inc.
> > > Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> > > Copyright (C) 1999-2006 Hewlett-Packard Co
> > > Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> > > Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> > > Copyright (C) 2005, 2011 NEC Corporation
> > > Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> > > Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> > > This program is free software, covered by the GNU General Public
> > > License,
> > > and you are welcome to change it and/or distribute copies of it under
> > > certain conditions. Enter "help copying" to see the
conditions.
> > > This program has absolutely no warranty. Enter "help
warranty" for
> > > details.
> > >
> > > GNU gdb (GDB) 7.6
> > > Copyright (C) 2013 Free Software Foundation, Inc.
> > > License GPLv3+: GNU GPL version 3 or later
> > > <http://gnu.org/licenses/gpl.html>
> > > This is free software: you are free to change and redistribute it.
> > > There is NO WARRANTY, to the extent permitted by law. Type "show
> > > copying"
> > > and "show warranty" for details.
> > > This GDB was configured as "--host=x86_64-unknown-linux-gnu
> > > --target=arm-elf-linux"...
> > >
> > > Segmentation fault (core dumped)
> > > $
> > >
> > > I haven't debugged it other than determining that the "note"
it looks to
> > > have found the single note OK, but then upon continuation the next time
> > > through the loop, the "note" pointer is valid at line 597, but
your
> > > function sets it back to NULL, and therefore it craps out at line 622:
> > >
> > > 597 note = (Elf32_Nhdr *)buf;
> > > 598 p = buf + sizeof(Elf32_Nhdr);
> > > 599
> > > 600 /*
> > > 601 * dumpfiles created with qemu won't have
> > > crash_notes, but there will
> > > 602 * be elf notes.
> > > 603 */
> > > 604 if (note->n_namesz == 0 &&
(DISKDUMP_DUMPFILE()
> > > ||
> > > KDUMP_DUMPFILE())) {
> > > 605 if (DISKDUMP_DUMPFILE())
> > > 606 note =
> > > diskdump_get_prstatus_percpu(i);
> > > 607 else if (KDUMP_DUMPFILE())
> > > 608 note =
> > > netdump_get_prstatus_percpu(i);
> > > 609 if (note) {
> > > 610 /*
> > > 611 * SIZE(note_buf) accounts for a
> > > "final note", which is a
> > > 612 * trailing empty elf note
> > > header.
> > > 613 */
> > > 614 long notesz = SIZE(note_buf) -
> > > sizeof(Elf32_Nhdr);
> > > 615
> > > 616 if (sizeof(Elf32_Nhdr) +
> > > roundup(note->n_namesz, 4) +
> > > 617 note->n_descsz == notesz)
> > > 618 BCOPY((char *)note, buf,
> > > notesz);
> > > 619 }
> > > 620 }
> > > 621
> > > 622 if (note->n_type != NT_PRSTATUS) {
> > > 623 error(WARNING, "invalid note (n_type
!=
> > > NT_PRSTATUS)\n");
> > > 624 goto fail;
> > > 625 }
> > >
> > > Not sure how you want to handle that, probably just bail out the same way
> > > if note becomes NULL?
> >
> > If I add this to arm_get_crash_notes(), just after your new function:
> >
> > if (!note) {
> > error(WARNING, "cannot find NT_PRSTATUS note for cpu:
%d\n",
> > i);
> > continue;
> > }
> >
> > I get this:
> >
> > $ ./crash /usr/dumps/ARM/vmcore.pae* /usr/dumps/ARM/vmlinux.pae.gz
> >
> > crash 7.1.4rc15
> > Copyright (C) 2002-2014 Red Hat, Inc.
> > Copyright (C) 2004, 2005, 2006, 2010 IBM Corporation
> > Copyright (C) 1999-2006 Hewlett-Packard Co
> > Copyright (C) 2005, 2006, 2011, 2012 Fujitsu Limited
> > Copyright (C) 2006, 2007 VA Linux Systems Japan K.K.
> > Copyright (C) 2005, 2011 NEC Corporation
> > Copyright (C) 1999, 2002, 2007 Silicon Graphics, Inc.
> > Copyright (C) 1999, 2000, 2001, 2002 Mission Critical Linux, Inc.
> > This program is free software, covered by the GNU General Public License,
> > and you are welcome to change it and/or distribute copies of it under
> > certain conditions. Enter "help copying" to see the conditions.
> > This program has absolutely no warranty. Enter "help warranty"
for
> > details.
> >
> > GNU gdb (GDB) 7.6
> > Copyright (C) 2013 Free Software Foundation, Inc.
> > License GPLv3+: GNU GPL version 3 or later
> > <http://gnu.org/licenses/gpl.html>
> > This is free software: you are free to change and redistribute it.
> > There is NO WARRANTY, to the extent permitted by law. Type "show
> > copying"
> > and "show warranty" for details.
> > This GDB was configured as "--host=x86_64-unknown-linux-gnu
> > --target=arm-elf-linux"...
> >
> > WARNING: cannot find NT_PRSTATUS note for cpu: 1
> > WARNING: cannot find NT_PRSTATUS note for cpu: 2
> > WARNING: cannot find NT_PRSTATUS note for cpu: 3
> > WARNING: cannot find NT_PRSTATUS note for cpu: 4
> > KERNEL: /usr/dumps/ARM/vmlinux.pae.gz
> > DUMPFILE: /usr/dumps/ARM/vmcore.pae
> > CPUS: 5 [OFFLINE: 4]
> > DATE: Sun Jun 8 18:27:39 2014
> > UPTIME: 00:03:22
> > LOAD AVERAGE: 0.16, 0.16, 0.07
> > TASKS: 51
> > NODENAME: buildroot
> > RELEASE: 3.13.5
> > VERSION: #3 SMP Mon Jun 9 05:58:39 CST 2014
> > MACHINE: armv7l (unknown Mhz)
> > MEMORY: 256 MB
> > PANIC: "SysRq : Trigger a crash"
> > PID: 732
> > COMMAND: "sh"
> > TASK: 8bcead00 [THREAD_INFO: 8ad32000]
> > CPU: 0
> > STATE: TASK_RUNNING (SYSRQ)
> >
> > crash> bt -a
> > PID: 732 TASK: 8bcead00 CPU: 0 COMMAND: "sh"
> > #0 [<80265064>] (sysrq_handle_crash) from [<80265810>]
> > #1 [<80265810>] (__handle_sysrq) from [<80265928>]
> > #2 [<80265928>] (write_sysrq_trigger) from [<80112120>]
> > #3 [<80112120>] (proc_reg_write) from [<800c9840>]
> > #4 [<800c9840>] (vfs_write) from [<800c9be4>]
> > #5 [<800c9be4>] (sys_write) from [<8000e3e0>]
> > pc : [<76e9cfdc>] lr : [<0000f998>] psr: 600d0010
> > sp : 7eab862c ip : 00000000 fp : 000a82a4
> > r10: 00000020 r9 : 000a8294 r8 : 00000001
> > r7 : 00000004 r6 : 000a9bf0 r5 : 00000001 r4 : 000a7d88
> > r3 : 00000000 r2 : 00000002 r1 : 000a9bf0 r0 : 00000001
> > Flags: nZCv IRQs on FIQs on Mode USER_32 ISA ARM
> >
> > PID: 0 TASK: 8bc561c0 CPU: 1 COMMAND: "swapper/1"
> > bt: WARNING: cannot determine starting stack frame for task 8bc561c0
> >
> > PID: 0 TASK: 8bc56580 CPU: 2 COMMAND: "swapper/2"
> > bt: WARNING: cannot determine starting stack frame for task 8bc56580
> >
> > PID: 0 TASK: 8bc56940 CPU: 3 COMMAND: "swapper/3"
> > bt: WARNING: cannot determine starting stack frame for task 8bc56940
> >
> > PID: 0 TASK: 8bc56d00 CPU: 4 COMMAND: "swapper/4"
> > bt: WARNING: cannot determine starting stack frame for task 8bc56d00
> > crash>
> >
> > Note that if I did "goto fail" instead of "continue", I
lose the good
> > cpu 0 backtrace from the NT_PRSTATUS that your patch found, so doing it
> > this way is the best of both worlds.
>
> I agree that the 'if (!note) continue' that you added is a good idea to
> try and salvage this type of dump. It shouldn't happen with qemu generated
> dumps, but anything's possible when a kernel panics...
>
> Would you like me to spin a v4 with this condition added? Or, since it
> actually seems to be addressing a non-qemu-generated dump issue, then
> maybe you just want to submit it as a new patch on top of the qemu
> patch?
No, I'm cleaning it up now -- there are a couple other gotcha's w/respect
to backtracing, since we're now going to keep the ms->panic_task_regs[cpus]
array intact, even if there's no note in some of them. I'm going to make
sure it's zero-filled in the not-found case, and I have to make sure that
it's recognized on both architectures. I'll post the the updated additions
to arm.c and arm64.c so that you can do a quick test on a legitimate
QEMU dump.
Might be Monday, I'm not sure...
Sounds good to me :-)
Thanks,
drew
10:35 a.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
----- Original Message -----
> > Would you like me to spin a v4 with this condition added?
Or, since it
> > actually seems to be addressing a non-qemu-generated dump issue, then
> > maybe you just want to submit it as a new patch on top of the qemu
> > patch?
>
> No, I'm cleaning it up now -- there are a couple other gotcha's w/respect
> to backtracing, since we're now going to keep the ms->panic_task_regs[cpus]
> array intact, even if there's no note in some of them. I'm going to make
> sure it's zero-filled in the not-found case, and I have to make sure that
> it's recognized on both architectures. I'll post the the updated additions
> to arm.c and arm64.c so that you can do a quick test on a legitimate
> QEMU dump.
>
> Might be Monday, I'm not sure...
Sounds good to me :-)
Thanks,
drew
Drew,
Can you sanity-test the attached patch? If it works OK for you, I'll post it as-is.
Thanks,
Dave
11:47 a.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
On Mon, Nov 23, 2015 at 11:35:38AM -0500, Dave Anderson wrote:
----- Original Message -----
> > > Would you like me to spin a v4 with this condition added? Or, since it
> > > actually seems to be addressing a non-qemu-generated dump issue, then
> > > maybe you just want to submit it as a new patch on top of the qemu
> > > patch?
> >
> > No, I'm cleaning it up now -- there are a couple other gotcha's
w/respect
> > to backtracing, since we're now going to keep the
ms->panic_task_regs[cpus]
> > array intact, even if there's no note in some of them. I'm going to
make
> > sure it's zero-filled in the not-found case, and I have to make sure that
> > it's recognized on both architectures. I'll post the the updated
additions
> > to arm.c and arm64.c so that you can do a quick test on a legitimate
> > QEMU dump.
> >
> > Might be Monday, I'm not sure...
>
> Sounds good to me :-)
>
> Thanks,
> drew
Drew,
Can you sanity-test the attached patch? If it works OK for you, I'll post it as-is.
Works. I tested on both arm and aarch64. I still think we should commit
it as two patches though. The qemu one, with its commit message focused
on qemu, and then this one, with its commit message focused on handling
missing prstatus for some cpus for unknown reasons. Actually, I don't
think it should even be possible to be missing prstatus for some cpus in
the qemu case (qemu will always populate them all), so the "dumpfiles
created with qemu..." comment might need some tweaking as well, since the
else-continue is being added to its if-block.
Thanks,
drew
Thanks,
Dave
diff --git a/arm.c b/arm.c
index ffc7c06..114a6ae 100644
--- a/arm.c
+++ b/arm.c
@@ -580,8 +580,8 @@ arm_get_crash_notes(void)
buf = GETBUF(SIZE(note_buf));
- if (!(panic_task_regs = malloc(kt->cpus*sizeof(*panic_task_regs))))
- error(FATAL, "cannot malloc panic_task_regs space\n");
+ if (!(panic_task_regs = calloc((size_t)kt->cpus, sizeof(*panic_task_regs))))
+ error(FATAL, "cannot calloc panic_task_regs space\n");
for (i=0;i<kt->cpus;i++) {
@@ -597,6 +597,32 @@ arm_get_crash_notes(void)
note = (Elf32_Nhdr *)buf;
p = buf + sizeof(Elf32_Nhdr);
+ /*
+ * dumpfiles created with qemu won't have crash_notes, but there will
+ * be elf notes.
+ */
+ if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE())) {
+ if (DISKDUMP_DUMPFILE())
+ note = diskdump_get_prstatus_percpu(i);
+ else if (KDUMP_DUMPFILE())
+ note = netdump_get_prstatus_percpu(i);
+ if (note) {
+ /*
+ * SIZE(note_buf) accounts for a "final note", which is a
+ * trailing empty elf note header.
+ */
+ long notesz = SIZE(note_buf) - sizeof(Elf32_Nhdr);
+
+ if (sizeof(Elf32_Nhdr) + roundup(note->n_namesz, 4) +
+ note->n_descsz == notesz)
+ BCOPY((char *)note, buf, notesz);
+ } else {
+ error(WARNING,
+ "cannot find NT_PRSTATUS note for cpu: %d\n", i);
+ continue;
+ }
+ }
+
if (note->n_type != NT_PRSTATUS) {
error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
goto fail;
@@ -835,6 +861,9 @@ arm_back_trace(struct bt_info *bt)
static void
arm_back_trace_cmd(struct bt_info *bt)
{
+ if (bt->flags & BT_REGS_NOT_FOUND)
+ return;
+
if (kt->flags & DWARF_UNWIND)
unwind_backtrace(bt);
else
@@ -1301,12 +1330,13 @@ arm_get_dumpfile_stack_frame(struct bt_info *bt, ulong *nip,
ulong *ksp)
{
const struct machine_specific *ms = machdep->machspec;
- if (!ms->crash_task_regs)
+ if (!ms->crash_task_regs ||
+ (!ms->crash_task_regs[bt->tc->processor].ARM_pc &&
+ !ms->crash_task_regs[bt->tc->processor].ARM_sp)) {
+ bt->flags |= BT_REGS_NOT_FOUND;
return FALSE;
+ }
- if (!is_task_active(bt->task))
- return FALSE;
-
/*
* We got registers for panic task from crash_notes. Just return them.
*/
@@ -1339,11 +1369,9 @@ arm_get_stack_frame(struct bt_info *bt, ulong *pcp, ulong *spp)
else
ret = arm_get_frame(bt, &ip, &sp);
- if (!ret) {
+ if (!ret)
error(WARNING, "cannot determine starting stack frame for task %lx\n",
bt->task);
- return;
- }
if (pcp)
*pcp = ip;
diff --git a/arm64.c b/arm64.c
index 06b45c2..2304a4c 100644
--- a/arm64.c
+++ b/arm64.c
@@ -1181,6 +1181,7 @@ arm64_back_trace_cmd(struct bt_info *bt)
stackframe.pc = bt->hp->eip ?
bt->hp->eip : GET_STACK_ULONG(bt->hp->esp);
stackframe.sp = bt->hp->esp + 8;
+ bt->flags &= ~BT_REGS_NOT_FOUND;
} else {
stackframe.sp = bt->stkptr;
stackframe.pc = bt->instptr;
@@ -1197,6 +1198,9 @@ arm64_back_trace_cmd(struct bt_info *bt)
return;
}
+ if (bt->flags & BT_REGS_NOT_FOUND)
+ return;
+
if (!(bt->flags & BT_KDUMP_ADJUST)) {
if (bt->flags & BT_USER_SPACE)
goto complete_user;
@@ -1336,8 +1340,12 @@ arm64_get_dumpfile_stackframe(struct bt_info *bt, struct
arm64_stackframe *frame
struct machine_specific *ms = machdep->machspec;
struct arm64_pt_regs *ptregs;
- if (!ms->panic_task_regs)
+ if (!ms->panic_task_regs ||
+ (!ms->panic_task_regs[bt->tc->processor].sp &&
+ !ms->panic_task_regs[bt->tc->processor].pc)) {
+ bt->flags |= BT_REGS_NOT_FOUND;
return FALSE;
+ }
ptregs = &ms->panic_task_regs[bt->tc->processor];
frame->sp = ptregs->sp;
@@ -1371,19 +1379,17 @@ static void
arm64_get_stack_frame(struct bt_info *bt, ulong *pcp, ulong *spp)
{
int ret;
- struct arm64_stackframe stackframe;
+ struct arm64_stackframe stackframe = { 0 };
if (DUMPFILE() && is_task_active(bt->task))
ret = arm64_get_dumpfile_stackframe(bt, &stackframe);
else
ret = arm64_get_stackframe(bt, &stackframe);
- if (!ret) {
+ if (!ret)
error(WARNING,
"cannot determine starting stack frame for task %lx\n",
bt->task);
- return;
- }
bt->frameptr = stackframe.fp;
if (pcp)
@@ -1811,8 +1817,8 @@ arm64_get_crash_notes(void)
buf = GETBUF(SIZE(note_buf));
- if (!(ms->panic_task_regs = malloc(kt->cpus * sizeof(struct arm64_pt_regs))))
- error(FATAL, "cannot malloc panic_task_regs space\n");
+ if (!(ms->panic_task_regs = calloc((size_t)kt->cpus, sizeof(struct
arm64_pt_regs))))
+ error(FATAL, "cannot calloc panic_task_regs space\n");
for (i = 0; i < kt->cpus; i++) {
@@ -1828,6 +1834,32 @@ arm64_get_crash_notes(void)
note = (Elf64_Nhdr *)buf;
p = buf + sizeof(Elf64_Nhdr);
+ /*
+ * dumpfiles created with qemu won't have crash_notes, but there will
+ * be elf notes.
+ */
+ if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE())) {
+ if (DISKDUMP_DUMPFILE())
+ note = diskdump_get_prstatus_percpu(i);
+ else if (KDUMP_DUMPFILE())
+ note = netdump_get_prstatus_percpu(i);
+ if (note) {
+ /*
+ * SIZE(note_buf) accounts for a "final note", which is a
+ * trailing empty elf note header.
+ */
+ long notesz = SIZE(note_buf) - sizeof(Elf64_Nhdr);
+
+ if (sizeof(Elf64_Nhdr) + roundup(note->n_namesz, 4) +
+ note->n_descsz == notesz)
+ BCOPY((char *)note, buf, notesz);
+ } else {
+ error(WARNING,
+ "cannot find NT_PRSTATUS note for cpu: %d\n", i);
+ continue;
+ }
+ }
+
if (note->n_type != NT_PRSTATUS) {
error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
goto fail;
diff --git a/defs.h b/defs.h
index 8e96461..dc79b9b 100644
--- a/defs.h
+++ b/defs.h
@@ -5048,6 +5048,7 @@ ulong cpu_map_addr(const char *type);
#define BT_EFRAME_TARGET (0x800000000000ULL)
#define BT_CPUMASK (0x1000000000000ULL)
#define BT_SHOW_ALL_REGS (0x2000000000000ULL)
+#define BT_REGS_NOT_FOUND (0x4000000000000ULL)
#define BT_SYMBOL_OFFSET (BT_SYMBOLIC_ARGS)
#define BT_REF_HEXVAL (0x1)
@@ -5721,6 +5722,7 @@ void dump_registers_for_qemu_mem_dump(void);
void kdump_backup_region_init(void);
void display_regs_from_elf_notes(int, FILE *);
void display_ELF_note(int, int, void *, FILE *);
+void *netdump_get_prstatus_percpu(int);
#define PRSTATUS_NOTE (1)
#define QEMU_NOTE (2)
diff --git a/netdump.c b/netdump.c
index bfa818f..3ff49f0 100644
--- a/netdump.c
+++ b/netdump.c
@@ -2345,6 +2345,27 @@ dump_Elf64_Nhdr(Elf64_Off offset, int store)
return len;
}
+void *
+netdump_get_prstatus_percpu(int cpu)
+{
+ int online;
+
+ if ((cpu < 0) || (cpu >= nd->num_prstatus_notes))
+ return NULL;
+
+ /*
+ * If no cpu mapping was done, then there must be
+ * a one-to-one relationship between the number
+ * of online cpus and the number of notes.
+ */
+ if ((online = get_cpus_online()) &&
+ (online == kt->cpus) &&
+ (online != nd->num_prstatus_notes))
+ return NULL;
+
+ return nd->nt_prstatus_percpu[cpu];
+}
+
/*
* Send the request to the proper architecture hander.
*/
12:48 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
----- Original Message -----
On Mon, Nov 23, 2015 at 11:35:38AM -0500, Dave Anderson wrote:
>
>
> ----- Original Message -----
> > > > Would you like me to spin a v4 with this condition added? Or, since
> > > > it
> > > > actually seems to be addressing a non-qemu-generated dump issue,
then
> > > > maybe you just want to submit it as a new patch on top of the qemu
> > > > patch?
> > >
> > > No, I'm cleaning it up now -- there are a couple other gotcha's
> > > w/respect
> > > to backtracing, since we're now going to keep the
> > > ms->panic_task_regs[cpus]
> > > array intact, even if there's no note in some of them. I'm going
to
> > > make
> > > sure it's zero-filled in the not-found case, and I have to make sure
> > > that
> > > it's recognized on both architectures. I'll post the the updated
> > > additions
> > > to arm.c and arm64.c so that you can do a quick test on a legitimate
> > > QEMU dump.
> > >
> > > Might be Monday, I'm not sure...
> >
> > Sounds good to me :-)
> >
> > Thanks,
> > drew
>
> Drew,
>
> Can you sanity-test the attached patch? If it works OK for you, I'll post
> it as-is.
Works. I tested on both arm and aarch64. I still think we should commit
it as two patches though. The qemu one, with its commit message focused
on qemu, and then this one, with its commit message focused on handling
missing prstatus for some cpus for unknown reasons. Actually, I don't
think it should even be possible to be missing prstatus for some cpus in
the qemu case (qemu will always populate them all), so the "dumpfiles
created with qemu..." comment might need some tweaking as well, since the
else-continue is being added to its if-block.
Thanks,
drew
All right, I'll post your v3 patch as-is, and make a follow-up patch
with my additions (plus a tweak to the comment).
Thanks,
Dave
>
> Thanks,
> Dave
>
> diff --git a/arm.c b/arm.c
> index ffc7c06..114a6ae 100644
> --- a/arm.c
> +++ b/arm.c
> @@ -580,8 +580,8 @@ arm_get_crash_notes(void)
>
> buf = GETBUF(SIZE(note_buf));
>
> - if (!(panic_task_regs = malloc(kt->cpus*sizeof(*panic_task_regs))))
> - error(FATAL, "cannot malloc panic_task_regs space\n");
> + if (!(panic_task_regs = calloc((size_t)kt->cpus,
> sizeof(*panic_task_regs))))
> + error(FATAL, "cannot calloc panic_task_regs space\n");
>
> for (i=0;i<kt->cpus;i++) {
>
> @@ -597,6 +597,32 @@ arm_get_crash_notes(void)
> note = (Elf32_Nhdr *)buf;
> p = buf + sizeof(Elf32_Nhdr);
>
> + /*
> + * dumpfiles created with qemu won't have crash_notes, but there will
> + * be elf notes.
> + */
> + if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE()))
{
> + if (DISKDUMP_DUMPFILE())
> + note = diskdump_get_prstatus_percpu(i);
> + else if (KDUMP_DUMPFILE())
> + note = netdump_get_prstatus_percpu(i);
> + if (note) {
> + /*
> + * SIZE(note_buf) accounts for a "final note", which is a
> + * trailing empty elf note header.
> + */
> + long notesz = SIZE(note_buf) - sizeof(Elf32_Nhdr);
> +
> + if (sizeof(Elf32_Nhdr) + roundup(note->n_namesz, 4) +
> + note->n_descsz == notesz)
> + BCOPY((char *)note, buf, notesz);
> + } else {
> + error(WARNING,
> + "cannot find NT_PRSTATUS note for cpu: %d\n", i);
> + continue;
> + }
> + }
> +
> if (note->n_type != NT_PRSTATUS) {
> error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
> goto fail;
> @@ -835,6 +861,9 @@ arm_back_trace(struct bt_info *bt)
> static void
> arm_back_trace_cmd(struct bt_info *bt)
> {
> + if (bt->flags & BT_REGS_NOT_FOUND)
> + return;
> +
> if (kt->flags & DWARF_UNWIND)
> unwind_backtrace(bt);
> else
> @@ -1301,12 +1330,13 @@ arm_get_dumpfile_stack_frame(struct bt_info *bt,
> ulong *nip, ulong *ksp)
> {
> const struct machine_specific *ms = machdep->machspec;
>
> - if (!ms->crash_task_regs)
> + if (!ms->crash_task_regs ||
> + (!ms->crash_task_regs[bt->tc->processor].ARM_pc &&
> + !ms->crash_task_regs[bt->tc->processor].ARM_sp)) {
> + bt->flags |= BT_REGS_NOT_FOUND;
> return FALSE;
> + }
>
> - if (!is_task_active(bt->task))
> - return FALSE;
> -
> /*
> * We got registers for panic task from crash_notes. Just return them.
> */
> @@ -1339,11 +1369,9 @@ arm_get_stack_frame(struct bt_info *bt, ulong *pcp,
> ulong *spp)
> else
> ret = arm_get_frame(bt, &ip, &sp);
>
> - if (!ret) {
> + if (!ret)
> error(WARNING, "cannot determine starting stack frame for task %lx\n",
> bt->task);
> - return;
> - }
>
> if (pcp)
> *pcp = ip;
> diff --git a/arm64.c b/arm64.c
> index 06b45c2..2304a4c 100644
> --- a/arm64.c
> +++ b/arm64.c
> @@ -1181,6 +1181,7 @@ arm64_back_trace_cmd(struct bt_info *bt)
> stackframe.pc = bt->hp->eip ?
> bt->hp->eip : GET_STACK_ULONG(bt->hp->esp);
> stackframe.sp = bt->hp->esp + 8;
> + bt->flags &= ~BT_REGS_NOT_FOUND;
> } else {
> stackframe.sp = bt->stkptr;
> stackframe.pc = bt->instptr;
> @@ -1197,6 +1198,9 @@ arm64_back_trace_cmd(struct bt_info *bt)
> return;
> }
>
> + if (bt->flags & BT_REGS_NOT_FOUND)
> + return;
> +
> if (!(bt->flags & BT_KDUMP_ADJUST)) {
> if (bt->flags & BT_USER_SPACE)
> goto complete_user;
> @@ -1336,8 +1340,12 @@ arm64_get_dumpfile_stackframe(struct bt_info *bt,
> struct arm64_stackframe *frame
> struct machine_specific *ms = machdep->machspec;
> struct arm64_pt_regs *ptregs;
>
> - if (!ms->panic_task_regs)
> + if (!ms->panic_task_regs ||
> + (!ms->panic_task_regs[bt->tc->processor].sp &&
> + !ms->panic_task_regs[bt->tc->processor].pc)) {
> + bt->flags |= BT_REGS_NOT_FOUND;
> return FALSE;
> + }
>
> ptregs = &ms->panic_task_regs[bt->tc->processor];
> frame->sp = ptregs->sp;
> @@ -1371,19 +1379,17 @@ static void
> arm64_get_stack_frame(struct bt_info *bt, ulong *pcp, ulong *spp)
> {
> int ret;
> - struct arm64_stackframe stackframe;
> + struct arm64_stackframe stackframe = { 0 };
>
> if (DUMPFILE() && is_task_active(bt->task))
> ret = arm64_get_dumpfile_stackframe(bt, &stackframe);
> else
> ret = arm64_get_stackframe(bt, &stackframe);
>
> - if (!ret) {
> + if (!ret)
> error(WARNING,
> "cannot determine starting stack frame for task %lx\n",
> bt->task);
> - return;
> - }
>
> bt->frameptr = stackframe.fp;
> if (pcp)
> @@ -1811,8 +1817,8 @@ arm64_get_crash_notes(void)
>
> buf = GETBUF(SIZE(note_buf));
>
> - if (!(ms->panic_task_regs = malloc(kt->cpus * sizeof(struct
> arm64_pt_regs))))
> - error(FATAL, "cannot malloc panic_task_regs space\n");
> + if (!(ms->panic_task_regs = calloc((size_t)kt->cpus, sizeof(struct
> arm64_pt_regs))))
> + error(FATAL, "cannot calloc panic_task_regs space\n");
>
> for (i = 0; i < kt->cpus; i++) {
>
> @@ -1828,6 +1834,32 @@ arm64_get_crash_notes(void)
> note = (Elf64_Nhdr *)buf;
> p = buf + sizeof(Elf64_Nhdr);
>
> + /*
> + * dumpfiles created with qemu won't have crash_notes, but there will
> + * be elf notes.
> + */
> + if (note->n_namesz == 0 && (DISKDUMP_DUMPFILE() || KDUMP_DUMPFILE()))
{
> + if (DISKDUMP_DUMPFILE())
> + note = diskdump_get_prstatus_percpu(i);
> + else if (KDUMP_DUMPFILE())
> + note = netdump_get_prstatus_percpu(i);
> + if (note) {
> + /*
> + * SIZE(note_buf) accounts for a "final note", which is a
> + * trailing empty elf note header.
> + */
> + long notesz = SIZE(note_buf) - sizeof(Elf64_Nhdr);
> +
> + if (sizeof(Elf64_Nhdr) + roundup(note->n_namesz, 4) +
> + note->n_descsz == notesz)
> + BCOPY((char *)note, buf, notesz);
> + } else {
> + error(WARNING,
> + "cannot find NT_PRSTATUS note for cpu: %d\n", i);
> + continue;
> + }
> + }
> +
> if (note->n_type != NT_PRSTATUS) {
> error(WARNING, "invalid note (n_type != NT_PRSTATUS)\n");
> goto fail;
> diff --git a/defs.h b/defs.h
> index 8e96461..dc79b9b 100644
> --- a/defs.h
> +++ b/defs.h
> @@ -5048,6 +5048,7 @@ ulong cpu_map_addr(const char *type);
> #define BT_EFRAME_TARGET (0x800000000000ULL)
> #define BT_CPUMASK (0x1000000000000ULL)
> #define BT_SHOW_ALL_REGS (0x2000000000000ULL)
> +#define BT_REGS_NOT_FOUND (0x4000000000000ULL)
> #define BT_SYMBOL_OFFSET (BT_SYMBOLIC_ARGS)
>
> #define BT_REF_HEXVAL (0x1)
> @@ -5721,6 +5722,7 @@ void dump_registers_for_qemu_mem_dump(void);
> void kdump_backup_region_init(void);
> void display_regs_from_elf_notes(int, FILE *);
> void display_ELF_note(int, int, void *, FILE *);
> +void *netdump_get_prstatus_percpu(int);
> #define PRSTATUS_NOTE (1)
> #define QEMU_NOTE (2)
>
> diff --git a/netdump.c b/netdump.c
> index bfa818f..3ff49f0 100644
> --- a/netdump.c
> +++ b/netdump.c
> @@ -2345,6 +2345,27 @@ dump_Elf64_Nhdr(Elf64_Off offset, int store)
> return len;
> }
>
> +void *
> +netdump_get_prstatus_percpu(int cpu)
> +{
> + int online;
> +
> + if ((cpu < 0) || (cpu >= nd->num_prstatus_notes))
> + return NULL;
> +
> + /*
> + * If no cpu mapping was done, then there must be
> + * a one-to-one relationship between the number
> + * of online cpus and the number of notes.
> + */
> + if ((online = get_cpus_online()) &&
> + (online == kt->cpus) &&
> + (online != nd->num_prstatus_notes))
> + return NULL;
> +
> + return nd->nt_prstatus_percpu[cpu];
> +}
> +
> /*
> * Send the request to the proper architecture hander.
> */
1:44 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
----- Original Message -----
> > Drew,
> >
> > Can you sanity-test the attached patch? If it works OK for you, I'll post
> > it as-is.
>
> Works. I tested on both arm and aarch64. I still think we should commit
> it as two patches though. The qemu one, with its commit message focused
> on qemu, and then this one, with its commit message focused on handling
> missing prstatus for some cpus for unknown reasons. Actually, I don't
> think it should even be possible to be missing prstatus for some cpus in
> the qemu case (qemu will always populate them all), so the "dumpfiles
> created with qemu..." comment might need some tweaking as well, since the
> else-continue is being added to its if-block.
>
> Thanks,
> drew
All right, I'll post your v3 patch as-is, and make a follow-up patch
with my additions (plus a tweak to the comment).
Thanks,
Your v3 patch, and then my subsequent update, are queued for crash-7.1.4:
https://github.com/crash-utility/crash/commit/b2d8f20407dd575356f028de9f7...
https://github.com/crash-utility/crash/commit/d963c2c15278b72bab23bacdf3f...
Thanks for your help with this,
Dave
Thanks,
Dave
1:57 p.m.
New subject: [PATCH v3 2/2] arm/arm64: read elf notes for qemu generated cores
On Mon, Nov 23, 2015 at 02:44:21PM -0500, Dave Anderson wrote:
----- Original Message -----
> > > Drew,
> > >
> > > Can you sanity-test the attached patch? If it works OK for you, I'll
post
> > > it as-is.
> >
> > Works. I tested on both arm and aarch64. I still think we should commit
> > it as two patches though. The qemu one, with its commit message focused
> > on qemu, and then this one, with its commit message focused on handling
> > missing prstatus for some cpus for unknown reasons. Actually, I don't
> > think it should even be possible to be missing prstatus for some cpus in
> > the qemu case (qemu will always populate them all), so the "dumpfiles
> > created with qemu..." comment might need some tweaking as well, since the
> > else-continue is being added to its if-block.
> >
> > Thanks,
> > drew
>
> All right, I'll post your v3 patch as-is, and make a follow-up patch
> with my additions (plus a tweak to the comment).
>
> Thanks,
Your v3 patch, and then my subsequent update, are queued for crash-7.1.4:
https://github.com/crash-utility/crash/commit/b2d8f20407dd575356f028de9f7...
https://github.com/crash-utility/crash/commit/d963c2c15278b72bab23bacdf3f...
Thanks for your help with this,
Dave
Thank you!
drew
Thanks,
Dave
3220
days inactive
3223
days old
devel@lists.crash-utility.osci.io
13 comments
2 participants
participants (2)
-
Andrew Jones -
Dave Anderson