Release 1.6.4-9
Resolves: RHEL-73070 Signed-off-by: Tao Liu <ltao@redhat.com>
This commit is contained in:
Tao Liu
parent
62adc343bb
commit
006156ca46
@ -0,0 +1,47 @@
|
||||
From 62b8d5005eaa9014467db85cfe268e65c6679f4c Mon Sep 17 00:00:00 2001
|
||||
From: Tao Liu <ltao@redhat.com>
|
||||
Date: Mon, 4 Nov 2024 17:11:19 +1300
|
||||
Subject: [PATCH 1/2] gcore: update set_context with upstream counterpart
|
||||
|
||||
With the introduction of upstream commit "Preparing for gdb stack unwind
|
||||
support" [1], the function set_context() is added by a 3rd parameter. Without
|
||||
this patch, the compiliation of gcore will fail.
|
||||
|
||||
The 3rd parameter of set_context() is used to sync the context of crash and
|
||||
gdb, so gdb can hold the the value of registers of current crash's task
|
||||
context, and gdb can output the stack unwinding for current task.
|
||||
|
||||
This have nothing to do with gcore, so simply set the 3rd parameter as FALSE.
|
||||
|
||||
[1]: https://github.com/lian-bo/crash/commit/d75d15d31b92f8882ccb15c960665e2c8a8d1c28
|
||||
|
||||
Signed-off-by: Tao Liu <ltao@redhat.com>
|
||||
---
|
||||
src/gcore.c | 4 ++--
|
||||
1 file changed, 2 insertions(+), 2 deletions(-)
|
||||
|
||||
diff --git a/src/gcore.c b/src/gcore.c
|
||||
index 47a9c0d..a79f69b 100644
|
||||
--- a/src/gcore.c
|
||||
+++ b/src/gcore.c
|
||||
@@ -306,7 +306,7 @@ static void do_gcore(char *arg)
|
||||
|
||||
if (tc != CURRENT_CONTEXT()) {
|
||||
gcore->orig_task = CURRENT_TASK();
|
||||
- (void) set_context(tc->task, NO_PID);
|
||||
+ (void) set_context(tc->task, NO_PID, FALSE);
|
||||
}
|
||||
|
||||
snprintf(gcore->corename, CORENAME_MAX_SIZE + 1, "core.%lu.%s",
|
||||
@@ -340,7 +340,7 @@ static void do_gcore(char *arg)
|
||||
}
|
||||
|
||||
if (gcore->orig_task)
|
||||
- (void)set_context(gcore->orig_task, NO_PID);
|
||||
+ (void)set_context(gcore->orig_task, NO_PID, FALSE);
|
||||
|
||||
}
|
||||
|
||||
--
|
||||
2.47.0
|
||||
|
||||
@ -0,0 +1,66 @@
|
||||
From b3af265aa60251f282145bb0eee3b83d8a15ebb7 Mon Sep 17 00:00:00 2001
|
||||
From: HATAYAMA Daisuke <d.hatayama@fujitsu.com>
|
||||
Date: Mon, 18 Sep 2023 20:45:11 -0400
|
||||
Subject: [PATCH] x86: fix extend gcore.so taking much time like more than 1
|
||||
hour when huge thread group is present
|
||||
|
||||
extend gcore.so command takes lengthy time such as more than 1 hour
|
||||
when huge thread group is present.
|
||||
|
||||
The root cause of this issue is that we never take into account time
|
||||
complexity of function gcore_arch_vsyscall_has_vm_alwaysdump_flag()
|
||||
sufficiently, which iterates the task list and vma lists of each task
|
||||
to detect existence of VM_ALWAYS flag in a given crash dump and so has
|
||||
O(nm) where the n is the length of the task list and the m is the
|
||||
total number of vma among all tasks.
|
||||
|
||||
To fix this issue, skip this process for the kernels whose version is
|
||||
equal to or larger than v3.4.0.
|
||||
|
||||
Originally, this process was implemented to detect existence of
|
||||
VM_ALWAYS flag that was removed at the commit
|
||||
909af768e88867016f427264ae39d27a57b6a8ed (coredump: remove
|
||||
VM_ALWAYSDUMP flag) included at Linux kernel v3.4. However, the base
|
||||
kernel version of RHEL7.0 GA is 3.10.0-123.el7. Hence, the kernels for
|
||||
RHEL7 and later have included the commit. For RHEL7 and later, it's
|
||||
sufficient to conclude absence of the flag by checking kernel version
|
||||
of a given crash dump.
|
||||
|
||||
The issue could still remain on RHEL6 and older, but we think it's
|
||||
acceptable because today we rarely handle crash dump files of RHEL6
|
||||
and older.
|
||||
|
||||
Signed-off-by: HATAYAMA Daisuke <d.hatayama@fujitsu.com>
|
||||
---
|
||||
src/libgcore/gcore_x86.c | 15 +++++++++++++++
|
||||
1 file changed, 15 insertions(+)
|
||||
|
||||
diff --git a/src/libgcore/gcore_x86.c b/src/libgcore/gcore_x86.c
|
||||
index 8878f79..3ddf510 100644
|
||||
--- a/src/libgcore/gcore_x86.c
|
||||
+++ b/src/libgcore/gcore_x86.c
|
||||
@@ -2502,6 +2502,21 @@ int gcore_arch_vsyscall_has_vm_alwaysdump_flag(void)
|
||||
struct task_context *tc;
|
||||
int i;
|
||||
|
||||
+ /*
|
||||
+ * The commit 909af768e88867016f427264ae39d27a57b6a8ed
|
||||
+ * (coredump: remove VM_ALWAYSDUMP flag) of the Linux kernel
|
||||
+ * was merged at v3.4. We can say VM_ALWAYSDUMP flag doesn't
|
||||
+ * exist for the version 3.4.0 or later. The base kernel
|
||||
+ * version of RHEL7.0 GA is 3.10.0-123.el7. Hence, this
|
||||
+ * condition is expected to match RHEL7 and later RHEL major
|
||||
+ * versions. On the other hand, the commit
|
||||
+ * 909af768e88867016f427264ae39d27a57b6a8ed was backported
|
||||
+ * into the kernel package in RHEL6 and the exploring code is
|
||||
+ * still needed.
|
||||
+ */
|
||||
+ if (THIS_KERNEL_VERSION >= LINUX(3, 4, 0))
|
||||
+ return FALSE;
|
||||
+
|
||||
/*
|
||||
* For simplicity, consider that VM_ALWAYSDUMP was already not
|
||||
* present when maple tree was introduced.
|
||||
--
|
||||
2.47.0
|
||||
|
||||
@ -0,0 +1,85 @@
|
||||
From e03ff7341a9a624a46cea4e60ee097606b1687a8 Mon Sep 17 00:00:00 2001
|
||||
From: HATAYAMA Daisuke <d.hatayama@fujitsu.com>
|
||||
Date: Sat, 23 Nov 2024 17:30:41 +0900
|
||||
Subject: [PATCH] x86: fix the issue that core files for 64-bit tasks are
|
||||
generated in the 32-bit format
|
||||
|
||||
Core files for 64-bit tasks are falsely generated in the 32-bit
|
||||
formats.
|
||||
|
||||
The root cause of this issue is that location of saving register
|
||||
values on kernel stacks is changed by the kernel commit
|
||||
65c9cc9e2c14602d98f1ca61c51ac954e9529303 (x86/fred: Reserve space for
|
||||
the FRED stack frame) in Linux kernel 6.9-rc1 to add some reserved
|
||||
area for FRED feature and gcore doesn't take this into
|
||||
consideration. Hence, the value of CS register retrieved from kernel
|
||||
stacks based on the existing logic is wrong and thus checking
|
||||
execution mode of a given task based on the value also works wrong.
|
||||
|
||||
To fix this issue, let's take the FRED feature into account. If FRED
|
||||
data structure is defined, retrieve register values from the address
|
||||
next to the reserved for the FRED feature.
|
||||
|
||||
This logic is borrowed from the commit
|
||||
48764a14bc5856f0b0bb30685336c68b832154fc (x86_64: fix for adding
|
||||
top_of_kernel_stack_padding for kernel stack) and the commit
|
||||
196c4b79c13d1c0e6d7b21c8321eca07d3838d6a (X86 64: fix a regression
|
||||
issue about kernel stack padding) of crash utility.
|
||||
---
|
||||
src/libgcore/gcore_x86.c | 9 ++++++---
|
||||
1 file changed, 6 insertions(+), 3 deletions(-)
|
||||
|
||||
diff --git a/src/libgcore/gcore_x86.c b/src/libgcore/gcore_x86.c
|
||||
index 3ddf510..4f43956 100644
|
||||
--- a/src/libgcore/gcore_x86.c
|
||||
+++ b/src/libgcore/gcore_x86.c
|
||||
@@ -514,10 +514,11 @@ convert_from_fxsr(struct user_i387_ia32_struct *env, struct task_context *target
|
||||
char *pt_regs_buf;
|
||||
uint16_t ds;
|
||||
struct machine_specific *ms = machdep->machspec;
|
||||
+ ulong padding_size = VALID_SIZE(fred_frame) > 0 ? (2 * 8) : 0;
|
||||
|
||||
pt_regs_buf = GETBUF(SIZE(pt_regs));
|
||||
|
||||
- readmem(machdep->get_stacktop(target->task) - SIZE(pt_regs),
|
||||
+ readmem(machdep->get_stacktop(target->task) - padding_size - SIZE(pt_regs),
|
||||
KVADDR, pt_regs_buf, SIZE(pt_regs),
|
||||
"convert_from_fxsr: regs",
|
||||
gcore_verbose_error_handle());
|
||||
@@ -1837,6 +1838,7 @@ static int genregs_get(struct task_context *target,
|
||||
struct user_regs_struct active_regs;
|
||||
const int active = is_task_active(target->task);
|
||||
struct machine_specific *ms = machdep->machspec;
|
||||
+ ulong padding_size = VALID_SIZE(fred_frame) > 0 ? (2 * 8) : 0;
|
||||
|
||||
BZERO(regs, sizeof(*regs));
|
||||
|
||||
@@ -1854,7 +1856,7 @@ static int genregs_get(struct task_context *target,
|
||||
*/
|
||||
pt_regs_buf = GETBUF(SIZE(pt_regs));
|
||||
|
||||
- readmem(machdep->get_stacktop(target->task) - SIZE(pt_regs), KVADDR,
|
||||
+ readmem(machdep->get_stacktop(target->task) - padding_size - SIZE(pt_regs), KVADDR,
|
||||
pt_regs_buf, SIZE(pt_regs), "genregs_get: pt_regs",
|
||||
gcore_verbose_error_handle());
|
||||
|
||||
@@ -2188,6 +2190,7 @@ static int genregs_get32(struct task_context *target,
|
||||
struct user_regs_struct *regs = (struct user_regs_struct *)buf;
|
||||
char *pt_regs_buf;
|
||||
ulonglong pt_regs_addr;
|
||||
+ ulong padding_size = VALID_SIZE(fred_frame) > 0 ? (2 * 8) : 0;
|
||||
|
||||
if (is_task_active(target->task) && KVMDUMP_DUMPFILE()) {
|
||||
get_regs_from_kvmdump_notes(target, regs);
|
||||
@@ -2198,7 +2201,7 @@ static int genregs_get32(struct task_context *target,
|
||||
|
||||
pt_regs_buf = GETBUF(SIZE(pt_regs));
|
||||
|
||||
- pt_regs_addr = machdep->get_stacktop(target->task) - SIZE(pt_regs);
|
||||
+ pt_regs_addr = machdep->get_stacktop(target->task) - padding_size - SIZE(pt_regs);
|
||||
|
||||
/*
|
||||
* The commit 07b047fc2466249aff7cdb23fa0b0955a7a00d48
|
||||
--
|
||||
2.47.0
|
||||
|
||||
@ -3,7 +3,7 @@
|
||||
Summary: Gcore extension module for the crash utility
|
||||
Name: crash-gcore-command
|
||||
Version: 1.6.4
|
||||
Release: 8%{?dist}
|
||||
Release: 9%{?dist}
|
||||
License: GPL-2.0-only
|
||||
Source0: https://github.com/fujitsu/crash-gcore/archive/v%{version}/%{name}-%{version}.tar.gz
|
||||
URL: https://github.com/fujitsu/crash-gcore
|
||||
@ -14,6 +14,9 @@ BuildRequires: gcc
|
||||
Requires: crash >= 5.1.5
|
||||
|
||||
Patch0: crash-gcore-1.6.4-coredump-fix-building-failure-due-to-undefined-macro.patch
|
||||
Patch1: 0001-x86-fix-extend-gcore.so-taking-much-time-like-more-t.patch
|
||||
Patch2: 0001-gcore-update-set_context-with-upstream-counterpart.patch
|
||||
Patch3: 0001-x86-fix-the-issue-that-core-files-for-64-bit-tasks-a.patch
|
||||
|
||||
%description
|
||||
Command for creating a core dump file of a user-space task that was
|
||||
@ -36,6 +39,9 @@ install -m 0755 -t %{buildroot}%{_libdir}/crash/extensions %{_builddir}/%{repona
|
||||
%license COPYING
|
||||
|
||||
%changelog
|
||||
* Mon May 5 2025 Tao Liu <ltao@redhat.com> - 1.6.4-9
|
||||
- rebase to latest upstream e03ff7341a
|
||||
|
||||
* Tue Oct 29 2024 Troy Dawson <tdawson@redhat.com> - 1.6.4-8
|
||||
- Bump release for October 2024 mass rebuild:
|
||||
Resolves: RHEL-64018
|
||||
|
||||
Loading…
Reference in New Issue
Block a user