[Crash-utility] Re: [PATCH] x86_64: Fix the bug of getting incorrect framesize

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Sep 16, 2024 at 3:46 PM <devel-request@xxxxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
Date: Mon, 16 Sep 2024 19:44:58 +1200
From: Tao Liu <ltao@xxxxxxxxxx>
Subject: [PATCH] x86_64: Fix the bug of getting
        incorrect framesize
To: devel@xxxxxxxxxxxxxxxxxxxxxxxxxxx
Cc: mycomplexlove@xxxxxxxxx
Message-ID: <20240916074458.105832-1-ltao@xxxxxxxxxx>
Content-Type: text/plain; charset="US-ASCII"; x-default=true

Previously, "retq" is used to determine the end of a function, so the end
of framesize calculation. However "ret" might be outputted by gdb rather
than "retq", as a result, the framesize is returned incorrectly, and bogus
stack trace will be outputted.

Without the patch:

   $ crash -d 3 vmcore vmlinux
   crash> bt
   0xffffffff92da7545 <copy_process+5>: push   %rbp     [framesize: 8]
   ...
   0xffffffff92da7561 <copy_process+33>:        sub    $0x238,%rsp      [framesize: 624]
   ...
   0xffffffff92da776a <copy_process+554>:       pop    %r15     [framesize: 8]
   0xffffffff92da776c <copy_process+556>:       pop    %rbp     [framesize: 0]
   0xffffffff92da776d <copy_process+557>:       ret

   crash> bt -D dump
   framesize_cache_entries:
      ...
      [  3]: ffffffff92dadcbd 0 CF (copy_process+26493)

   crash> bt
   ...
   #9  [ffff888263157bc0] copy_process at ffffffff92dadcbd
   #10 [ffff888263157d20] __mutex_init at ffffffff92ed8dd5
   #11 [ffff888263157d38] __alloc_file at ffffffff93458397
   #12 [ffff888263157d60] alloc_empty_file at ffffffff934585d2
   #13 [ffff888263157da8] __alloc_fd at ffffffff934b5ead
   #14 [ffff888263157e38] _do_fork at ffffffff92dae7a1
   #15 [ffff888263157f28] do_syscall_64 at ffffffff92c085f4

Stack #10 ~ #13 are bogus and misleading.

With the patch:
   ...
   0xffffffff92da776d <copy_process+557>:       ret     [framesize restored to: 624]

   crash> bt -D dump
      ...
      [  3]: ffffffff92dadcbd 624 CF (copy_process+26493)

   crash> bt
   ...
   #9  [ffff888263157bc0] copy_process at ffffffff92dadcbd
   #10 [ffff888263157e38] _do_fork at ffffffff92dae7a1
   #11 [ffff888263157f28] do_syscall_64 at ffffffff92c085f4

Signed-off-by: Tao Liu <ltao@xxxxxxxxxx>
---
 x86_64.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/x86_64.c b/x86_64.c
index 469d26b..7aa9430 100644
--- a/x86_64.c
+++ b/x86_64.c
@@ -8781,7 +8781,8 @@ x86_64_get_framesize(struct bt_info *bt, ulong textaddr, ulong rsp, char *stack_
                        if (CRASHDEBUG(2) || (bt->flags & BT_FRAMESIZE_DEBUG))
                                fprintf(fp, "%s\t[framesize: %d]\n",
                                        strip_linefeeds(buf2), framesize);
-               } else if (STRNEQ(arglist[instr], "retq")) {
+               } else if (STRNEQ(arglist[instr], "retq") ||
+                          STRNEQ(arglist[instr], "ret")) {

Thank you for the fix, Tao.

This looks good.  Applied:
https://github.com/crash-utility/crash/commit/0d2ad774532db3c4dad6cda05d51db74d0e3fa86


Lianbo 

                        if (!exception) {
                                framesize = max;
                                if (CRASHDEBUG(2) || (bt->flags & BT_FRAMESIZE_DEBUG))
--
2.40.1
--
Crash-utility mailing list -- devel@xxxxxxxxxxxxxxxxxxxxxxxxxxx
To unsubscribe send an email to devel-leave@xxxxxxxxxxxxxxxxxxxxxxxxxxx
https://${domain_name}/admin/lists/devel.lists.crash-utility.osci.io/
Contribution Guidelines: https://github.com/crash-utility/crash/wiki

[Index of Archives]     [Fedora Development]     [Fedora Desktop]     [Fedora SELinux]     [Yosemite News]     [KDE Users]     [Fedora Tools]

 

Powered by Linux