Skip to content

Commit

Permalink
x86: make traps on 'iret' be debuggable in user space
Browse files Browse the repository at this point in the history
This makes a trap on the 'iret' that returns us to user space
cause a nice clean SIGSEGV, instead of just a hard (and silent)
exit.

That way a debugger can actually try to see what happened, and
we also properly notify everybody who might be interested about
us being gone.

This loses the error code, but tells the debugger what happened
with ILL_BADSTK in the siginfo.
  • Loading branch information
Linus Torvalds committed Apr 29, 2005
1 parent c06fec5 commit a879cbb
Show file tree
Hide file tree
Showing 2 changed files with 4 additions and 5 deletions.
8 changes: 3 additions & 5 deletions arch/i386/kernel/entry.S
Original file line number Diff line number Diff line change
Expand Up @@ -260,11 +260,9 @@ restore_nocheck:
.section .fixup,"ax"
iret_exc:
sti
movl $__USER_DS, %edx
movl %edx, %ds
movl %edx, %es
movl $11,%eax
call do_exit
pushl $0 # no error code
pushl $do_iret_error
jmp error_code
.previous
.section __ex_table,"a"
.align 4
Expand Down
1 change: 1 addition & 0 deletions arch/i386/kernel/traps.c
Original file line number Diff line number Diff line change
Expand Up @@ -451,6 +451,7 @@ DO_ERROR(10, SIGSEGV, "invalid TSS", invalid_TSS)
DO_ERROR(11, SIGBUS, "segment not present", segment_not_present)
DO_ERROR(12, SIGBUS, "stack segment", stack_segment)
DO_ERROR_INFO(17, SIGBUS, "alignment check", alignment_check, BUS_ADRALN, 0)
DO_ERROR_INFO(32, SIGSEGV, "iret exception", iret_error, ILL_BADSTK, 0)

fastcall void do_general_protection(struct pt_regs * regs, long error_code)
{
Expand Down

0 comments on commit a879cbb

Please sign in to comment.