leandrof/linux
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
944a343861
linux/arch/sh/kernel
History
Matt Fleming 944a343861 sh: Don't continue unwinding across interrupts 
Unfortunately, due to poor DWARF info in current toolchains, unwinding
through interrutps cannot be done reliably. The problem is that the
DWARF info for function epilogues is wrong.

Take this standard epilogue sequence,

80003cc4:       e3 6f           mov     r14,r15
80003cc6:       26 4f           lds.l   @r15+,pr
80003cc8:       f6 6e           mov.l   @r15+,r14
						<---- interrupt here
80003cca:       f6 6b           mov.l   @r15+,r11
80003ccc:       f6 6a           mov.l   @r15+,r10
80003cce:       f6 69           mov.l   @r15+,r9
80003cd0:       0b 00           rts

If we take an interrupt at the highlighted point, the DWARF info will
bogusly claim that the return address can be found at some offset from
the frame pointer, even though the frame pointer was just restored. The
worst part is if the unwinder finds a text address at the bogus stack
address - unwinding will continue, for a bit, until it finally comes
across an unexpected address on the stack and blows up.

The only solution is to stop unwinding once we've calculated the
function that was executing when the interrupt occurred. This PC can be
easily calculated from pt_regs->pc.

Signed-off-by: Matt Fleming <matt@console-pimps.org>
Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2010-02-08 10:47:04 +09:00
..
cpu sh: Setup frame pointer in handle_exception path 2010-02-08 10:46:53 +09:00
vsyscall kbuild: rename ld-option to cc-ldoption 2009-09-20 12:27:42 +02:00
.gitignore
asm-offsets.c sh: Add R-standby sleep mode support 2009-10-30 14:38:45 +09:00
cpufreq.c sh: cpufreq: Include CPU id in info messages. 2009-09-15 09:26:04 +09:00
crash_dump.c
debugtraps.S sh: unwinder: Use a special bug flag for unwinder traps. 2009-08-22 05:28:25 +09:00
disassemble.c
dma-nommu.c sh: Revamp PCI DMA coherence Kconfig bits. 2009-10-27 10:35:02 +09:00
dumpstack.c sh: flag the default unwinder as reliable. 2009-08-15 01:09:03 +09:00
dwarf.c sh: Don't continue unwinding across interrupts 2010-02-08 10:47:04 +09:00
early_printk.c sh: Remove old early serial console code V2 2009-12-15 12:07:35 +09:00
entry-common.S sh: Correct the offset of the return address in ret_from_exception 2010-02-08 10:46:46 +09:00
ftrace.c sh: ftrace: Fix up syscall tracing build. 2009-12-11 15:25:41 +09:00
head_32.S sh: CONFIG_PMB doesn't mean the MMU is in 32bit mode 2009-10-10 21:51:23 +09:00
head_64.S sh: convert to use __HEAD and HEAD_TEXT macros. 2009-04-26 09:20:39 -07:00
idle.c sh: Only use bl bit toggling for sleeping idle. 2009-12-21 11:57:33 +09:00
init_task.c Use new __init_task_data macro in arch init_task.c files. 2009-09-21 06:27:08 +02:00
io_generic.c sh: Use the generic I/O port base for slowdown. 2009-11-12 16:36:26 +09:00
io_trapped.c sh: Only shout about fixing up unexpected unaligned accesses 2009-08-29 20:31:44 +09:00
io.c sh: Optimise memcpy_to/fromio for SH4 2009-08-24 17:35:07 +09:00
irq_32.c sh: Annotate irq functions with "notrace" 2009-10-27 07:37:10 +09:00
irq_64.c sh: Annotate irq functions with "notrace" 2009-10-27 07:37:10 +09:00
irq.c genirq: Convert irq_desc.lock to raw_spinlock 2009-12-14 23:55:33 +01:00
kgdb.c sh: Kill off kgdb's magical NMI debouncing. 2009-09-01 17:45:35 +09:00
kprobes.c
localtimer.c sh: Fix up clockevents broadcasting. 2009-08-19 18:00:03 +09:00
machine_kexec.c sh: fix kexec by removing check for old kexec-tools 2009-10-26 19:45:45 +09:00
machvec.c sh: Use the generic I/O port base for slowdown. 2009-11-12 16:36:26 +09:00
Makefile sh: Remove old early serial console code V2 2009-12-15 12:07:35 +09:00
module.c sh: Tidy up the dwarf module helpers. 2009-10-13 13:32:19 +09:00
nmi_debug.c sh: nmi_debug support. 2009-09-01 17:38:32 +09:00
perf_callchain.c sh: perf events: Preliminary callchain support. 2009-11-05 16:20:09 +09:00
perf_event.c sh: perf events: Fix up uninitialized variable warning. 2009-11-05 13:56:50 +09:00
process_32.c sh: Fix up the FPU emulation build. 2009-11-25 12:07:31 +09:00
process_64.c Split 'flush_old_exec' into two functions 2010-01-29 08:22:01 -08:00
ptrace_32.c sh: Wire up HAVE_SYSCALL_TRACEPOINTS. 2009-09-16 14:30:34 +09:00
ptrace_64.c sh: Fix up FPU build for SH5 2009-12-15 12:07:36 +09:00
relocate_kernel.S sh: add kexec jump support 2009-03-18 18:57:43 +09:00
return_address.c sh: Generalize CALLER_ADDRx support. 2009-10-13 13:10:14 +09:00
setup.c sh-sci: Extend sh-sci driver with early console V2 2009-12-15 12:06:29 +09:00
sh_bios.c
sh_ksyms_32.c sh: Clean up more superfluous symbol exports. 2009-10-27 11:51:19 +09:00
sh_ksyms_64.c sh: Clean up more superfluous symbol exports. 2009-10-27 11:51:19 +09:00
signal_32.c sh: TS_RESTORE_SIGMASK conversion. 2009-10-14 16:05:42 +09:00
signal_64.c sh: Fix up FPU build for SH5 2009-12-15 12:07:36 +09:00
smp.c Merge branch 'sh/stable-updates' 2009-10-14 15:53:08 +09:00
stacktrace.c sh: stacktrace: Properly terminate the trace entry buffer. 2009-08-15 01:11:37 +09:00
swsusp.c sh: hibernation support 2009-03-10 12:55:40 +09:00
sys_sh32.c headers: utsname.h redux 2009-09-23 18:13:10 -07:00
sys_sh64.c headers: utsname.h redux 2009-09-23 18:13:10 -07:00
sys_sh.c Unify sys_mmap* 2009-12-11 06:44:29 -05:00
syscalls_32.S sh: unwire sys_recvmmsg. 2010-01-19 17:00:06 +09:00
syscalls_64.S sh64: wire up sys_accept4. 2010-01-19 17:00:31 +09:00
time.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6 2009-09-18 09:43:09 -07:00
topology.c sh: Support SCHED_MC for SH-X3 multi-cores. 2009-10-16 18:00:02 +09:00
traps_32.c sh: Make the unaligned trap handler always obey notification levels. 2009-12-14 11:46:09 +09:00
traps_64.c sh: Fix up FPU build for SH5 2009-12-15 12:07:36 +09:00
traps.c sh64: Fix up the CONFIG_GENERIC_BUG=n build. 2009-11-12 16:39:47 +09:00
unwinder.c sh: unwinder: Use a special bug flag for unwinder traps. 2009-08-22 05:28:25 +09:00
vmlinux.lds.S Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6 2009-09-16 13:48:32 +09:00