Hi,hatayama
it looks that the commit c975008e61121ef8785622c3bc26964da8fe0deb of
crash utility
has the condition MEMBER_EXISTS("pt_regs", "stackframe").
This
condition can be used to judge before or after 4.7,This change is accompanied by the
variable stackframe.
I guess with this patch set, gcore would result in similar issue for
kernels before 4.7.
Before kernel 4.7,user_eframe_offset is "SIZE(pt_regs) + 16",after 4.7
user_eframe_offset is SIZE(pt_regs) ,Anderson's change can compatible with before and
after kernel 4.7,so gcore can use "user_eframe_offset" directly.
+ if (MEMBER_EXISTS("pt_regs", "stackframe"))
+ machdep->machspec->user_eframe_offset = SIZE(pt_regs);
+ else
+ machdep->machspec->user_eframe_offset = SIZE(pt_regs) + 16;
+
________________________________________
From: d.hatayama(a)fujitsu.com <d.hatayama(a)fujitsu.com>
Sent: Friday, November 13, 2020 14:20
To: 赵乾利
Cc: Dreyfus, Thirtsa (Mobileye); lijiang(a)redhat.com; crash-utility(a)redhat.com;
k-hagio-ab(a)nec.com
Subject: Re: [External Mail][ANNOUNCE] crash gcore command, version 1.6.0 is released
Hi,hatayama
Hi Zhao. Thanks for your prompt response.
I've noticed a issue that has not been fixed in this version,gcore unable to parse
register correctly in arm64,i found crash-utility is fixed,gcore-command can refer to
below change:
I see. I'll include fix for this issue into v1.6.1.
But I'm waiting for bug reports from other archs a few more weeks.
By the way,
commit c975008e61121ef8785622c3bc26964da8fe0deb
Author: Dave Anderson <anderson(a)redhat.com>
Date: Fri Sep 22 14:59:10 2017 -0400
Fix for the ARM64 "bt" command's display of the user mode exception
frame at the top of the stack in Linux 4.7 and later kernels.
Without the patch, the contents of the user mode exception frame are
invalid due to the miscalculation of the starting address of the
pt_regs structure on the kernel stack.
(anderson(a)redhat.com)
gcore patch:
commit 837182cc6589095c0d08f71f57953c50ad61cc19
Author: zhaoqianli <zhaoqianli(a)xiaomi.com>
Date: Thu Nov 12 19:41:01 2020 +0800
Fix register parsing error caused by miscalculation of the
starting address of the pt_regs structure on the kernel stack
Signed-off-by: zhaoqianli <zhaoqianli(a)xiaomi.com>
diff --git a/libgcore/gcore_arm64.c b/libgcore/gcore_arm64.c
index 3257389..c828fee 100644
--- a/libgcore/gcore_arm64.c
+++ b/libgcore/gcore_arm64.c
@@ -28,7 +28,7 @@ static int gpr_get(struct task_context *target,
BZERO(regs, sizeof(*regs));
- readmem(machdep->get_stacktop(target->task) - 16 - SIZE(pt_regs), KVADDR,
+ readmem(machdep->get_stacktop(target->task) -
machdep->machspec->user_eframe_offset, KVADDR,
it looks that the commit c975008e61121ef8785622c3bc26964da8fe0deb of crash utility
has the condition MEMBER_EXISTS("pt_regs", "stackframe").
@@ -1376,6 +1377,11 @@ arm64_stackframe_init(void)
MEMBER_OFFSET_INIT(elf_prstatus_pr_pid, "elf_prstatus",
"pr_pid");
MEMBER_OFFSET_INIT(elf_prstatus_pr_reg, "elf_prstatus",
"pr_reg");
+ if (MEMBER_EXISTS("pt_regs", "stackframe"))
+ machdep->machspec->user_eframe_offset = SIZE(pt_regs);
+ else
+ machdep->machspec->user_eframe_offset = SIZE(pt_regs) + 16;
+
machdep->machspec->__exception_text_start =
symbol_value("__exception_text_start");
machdep->machspec->__exception_text_end =
@
I guess with this patch set, gcore would result in similar issue for
kernels before 4.7.
Could you improve the patch so that gcore works well for kernls both before and later v4.7
kernel?
Thanks.
HATAYAMA, Daisuke
#/******本邮件及其附件含有小米公司的保密信息,仅限于发送给上面地址中列出的个人或群组。禁止任何其他人以任何形式使用(包括但不限于全部或部分地泄露、复制、或散发)本邮件中的信息。如果您错收了本邮件,请您立即电话或邮件通知发件人并删除本邮件!
This e-mail and its attachments contain confidential information from XIAOMI, which is
intended only for the person or entity whose address is listed above. Any use of the
information contained herein in any way (including, but not limited to, total or partial
disclosure, reproduction, or dissemination) by persons other than the intended
recipient(s) is prohibited. If you receive this e-mail in error, please notify the sender
by phone or email immediately and delete it!******/#