lkml.org 
[lkml]   [2005]   [Jun]   [16]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
Subject2.6.11.11 x86_64 gdb passes ERESTARTNOHAND to user process
I'm chasing a problem on linux 2.6.11.11 x86_64 where when you attach gdb,
select() returns ERESTARTNOHAND (errno 514) to the user process.
The same problem occurs whether it is a 32bit or 64bit process being debugged.

I'd be interested in any comments you may have.

How to reproduce:
- attach to a process with gdb, or stop the running process within gdb
- call a function, eg 'call sleep(1)'
- continue program execution

On i386, it seems as if the EIP is backed up 2 insn's to execute the
'jmp' insn that causes the system call to be restarted.

On x86_64, the EIP is not backed up, the system call is not restarted, and
the internal kernel errno is passed on to the user process.

Thanks.

:::: calling a function on x86_64 ::::
Program received signal SIGINT, Interrupt.
0xffffe405 in __kernel_vsyscall ()
1: x/i $pc 0xffffe405 <__kernel_vsyscall+5>: mov $0x2b,%ecx
(gdb) call doNothing()
(gdb) x/i $pc
0xffffe405 <__kernel_vsyscall+5>: mov $0x2b,%ecx
(gdb) x/8i 0xffffe400
0xffffe400 <__kernel_vsyscall>: push %ebp
0xffffe401 <__kernel_vsyscall+1>: mov %ecx,%ebp
0xffffe403 <__kernel_vsyscall+3>: syscall
0xffffe405 <__kernel_vsyscall+5>: mov $0x2b,%ecx
0xffffe40a <__kernel_vsyscall+10>: movl %ecx,%ss
0xffffe40c <__kernel_vsyscall+12>: mov %ebp,%ecx
0xffffe40e <__kernel_vsyscall+14>: pop %ebp
0xffffe40f <__kernel_vsyscall+15>: ret
(gdb) stepi
0xffffe40a in __kernel_vsyscall ()
1: x/i $pc 0xffffe40a <__kernel_vsyscall+10>: movl %ecx,%ss
(gdb) stepi
0xffffe40e in __kernel_vsyscall ()
1: x/i $pc 0xffffe40e <__kernel_vsyscall+14>: pop %ebp
:::: calling a function on i386 ::::
Program received signal SIGINT, Interrupt.
0xffffe410 in __kernel_vsyscall ()
1: x/i $pc 0xffffe410 <__kernel_vsyscall+16>: pop %ebp
(gdb) x/8i 0xffffe400
0xffffe400 <__kernel_vsyscall>: push %ecx
0xffffe401 <__kernel_vsyscall+1>: push %edx
0xffffe402 <__kernel_vsyscall+2>: push %ebp
0xffffe403 <__kernel_vsyscall+3>: mov %esp,%ebp
0xffffe405 <__kernel_vsyscall+5>: sysenter
0xffffe407 <__kernel_vsyscall+7>: nop
0xffffe408 <__kernel_vsyscall+8>: nop
0xffffe409 <__kernel_vsyscall+9>: nop
0xffffe40a <__kernel_vsyscall+10>: nop
0xffffe40b <__kernel_vsyscall+11>: nop
0xffffe40c <__kernel_vsyscall+12>: nop
0xffffe40d <__kernel_vsyscall+13>: nop
0xffffe40e <__kernel_vsyscall+14>: jmp 0xffffe403 <__kernel_vsyscall+3>
0xffffe410 <__kernel_vsyscall+16>: pop %ebp
0xffffe411 <__kernel_vsyscall+17>: pop %edx
0xffffe412 <__kernel_vsyscall+18>: pop %ecx
0xffffe413 <__kernel_vsyscall+19>: ret
(gdb) call doNothing()
(gdb) stepi
0xffffe403 in __kernel_vsyscall ()
1: x/i $pc 0xffffe403 <__kernel_vsyscall+3>: mov %esp,%ebp
(gdb) c
Continuing.
Program received signal SIGINT, Interrupt.
0xffffe410 in __kernel_vsyscall ()
1: x/i $pc 0xffffe410 <__kernel_vsyscall+16>: pop %ebp
(gdb) call doNothing()
(gdb) x/i $pc
0xffffe410 <__kernel_vsyscall+16>: pop %ebp
(gdb) stepi
0xffffe403 in __kernel_vsyscall ()
1: x/i $pc 0xffffe403 <__kernel_vsyscall+3>: mov %esp,%ebp
--
Rich Coe richard.coe@med.ge.com
General Electric Healthcare Technologies
Global Software Platforms, Computer Technology Team
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-06-16 15:20    [from the cache]
©2003-2011 Jasper Spaans