segmentation fault on power system

Hey there,

When trying to run fio on one of our power system, segmentation fault

Quoted 4 lines · Han Pingtian

Hey there,

When trying to run fio on one of our power system, segmentation fault

Can you give us the kind of cpu you are using ? (/proc/cpuinfo) I'm not used with PPC but maybe your processor doesn't support ATBU call on mfspr.

Quoted 5 lines · Erwan Velu

Quoted 4 lines · Han Pingtian

Hey there,

When trying to run fio on one of our power system, segmentation fault

Quoted 3 lines

Can you give us the kind of cpu you are using ? (/proc/cpuinfo) I'm not used with PPC but maybe your processor doesn't support ATBU call on mfspr.

That is definitely the problem, so CPU info would help. In the mean time, you can use clocksource=clock_gettime to get rid of the illegal instruction.

Quoted 6 lines · Jens Axboe

Quoted 5 lines · Erwan Velu

Quoted 4 lines · Han Pingtian

Hey there,

When trying to run fio on one of our power system, segmentation fault

Quoted 7 lines

Quoted 3 lines

Can you give us the kind of cpu you are using ? (/proc/cpuinfo) I'm not used with PPC but maybe your processor doesn't support ATBU call on mfspr.

That is definitely the problem, so CPU info would help. In the mean time, you can use clocksource=clock_gettime to get rid of the illegal instruction.

This is the contents of /proc/cpuinfo:

processor : 0
cpu : POWER7 (architected), altivec supported
clock : 3550.000000MHz
revision : 2.3 (pvr 003f 0203)

processor : 1
cpu : POWER7 (architected), altivec supported
clock : 3550.000000MHz
revision : 2.3 (pvr 003f 0203)

processor : 2
cpu : POWER7 (architected), altivec supported
clock : 3550.000000MHz
revision : 2.3 (pvr 003f 0203)

processor : 3
cpu : POWER7 (architected), altivec supported
clock : 3550.000000MHz
revision : 2.3 (pvr 003f 0203)

timebase : 512000000
platform : pSeries
model : IBM,8231-E2C
machine : CHRP IBM,8231-E2C

But 'clocksource=clock_gettime' doesn't fix the fault. I can also get the same two cores.

If I changed the code like this:

================================================================================

arch/arch-ppc.h +6 -6

index 65e6b74..30c315c 100644
@@ -67,15 +67,15 @@ static inline unsigned long long get_cpu_clock(void)
        unsigned long long ret;
 
        do {
-               if (arch_flags & ARCH_FLAG_1) {
-                       tbu0 = mfspr(SPRN_ATBU);
-                       tbl = mfspr(SPRN_ATBL);
-                       tbu1 = mfspr(SPRN_ATBU);
-               } else {
+               //if (arch_flags & ARCH_FLAG_1) {
+               //      tbu0 = mfspr(SPRN_ATBU);
+               //      tbl = mfspr(SPRN_ATBL);
+               //      tbu1 = mfspr(SPRN_ATBU);
+               //} else {
                        tbu0 = mfspr(SPRN_TBRU);
                        tbl = mfspr(SPRN_TBRL);
                        tbu1 = mfspr(SPRN_TBRU);
-               }
+               //}
        } while (tbu0 != tbu1);
 
        ret = (((unsigned long long)tbu0) << 32) | tbl;

================================================================================

then only one core dumpped which has this backtrace:

================================================================================
Core was generated by `./fio/fio --debug=parse fio-jobs/randomw.fio '. Program terminated with signal 11, Segmentation fault.
#0 0x000000001006288c in init_disk_util (td=0xfff9a730000) at diskutil.c:481
481 if (!td->o.do_disk_util ||
(gdb) bt

#0  0x000000001006288c in init_disk_util (td=0xfff9a730000) at diskutil.c:481
#1  0x0000000010050d30 in run_threads () at backend.c:1691
#2  0x000000001005159c in fio_backend () at backend.c:1911
#3  0x00000000100669a4 in main (argc=<value optimized out>, argv=0xfffd20c9ec8, envp=<value optimized out>)
    at fio.c:50

(gdb) p td
$1 = (struct thread_data *) 0xfff9a730000
(gdb) p threads
$2 = (struct thread_data *) 0xfff9a730000
(gdb) p td->io_ops
$3 = (struct ioengine_ops *) 0x0
(gdb) p threads->io_ops
$4 = (struct ioengine_ops *) 0x1000e262720
(gdb)
================================================================================

--
To unsubscribe from this list: send the line "unsubscribe fio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

I think I have found what's wrong. After 'make distclean' then 'make', the problem of segfault has gone. There were some object files compiled on another machine. The recompilation fixed it. But looks like the ATBU call dosen't work on power7 system.

Quoted 4 lines · Han Pingtian

I think I have found what's wrong. After 'make distclean' then 'make', the problem of segfault has gone. There were some object files compiled on another machine. The recompilation fixed it. But looks like the ATBU call dosen't work on power7 system.

So the recompilation fixed the segfault, but not the illegal instruction due to ATBU?

Quoted 8 lines · Jens Axboe

Quoted 4 lines · Han Pingtian

I think I have found what's wrong. After 'make distclean' then 'make', the problem of segfault has gone. There were some object files compiled on another machine. The recompilation fixed it. But looks like the ATBU call dosen't work on power7 system.

So the recompilation fixed the segfault, but not the illegal instruction due to ATBU?

Yes. If we set unlimit core dump, fio will generate a core with backtrace like this one:

#0 mfspr (envp=<value optimized out>) at arch/arch-ppc.h:55
55 asm volatile("mfspr %0,%1": "=r" (val) : "K" (reg));
Missing separate debuginfos, use: debuginfo-install glibc-2.12-1.107.el6.ppc64 libaio-0.3.107-10.el6.ppc64 (gdb) bt

#0  mfspr (envp=<value optimized out>) at arch/arch-ppc.h:55
#1  get_cpu_clock (envp=<value optimized out>) at arch/arch-ppc.h:71
#2  atb_child (envp=<value optimized out>) at arch/arch-ppc.h:88
#3  atb_clocktest (envp=<value optimized out>) at arch/arch-ppc.h:98
#4  arch_init (envp=<value optimized out>) at arch/arch-ppc.h:116
#5  initialize_fio (envp=<value optimized out>) at libfio.c:248
#6  0x0000000010067294 in main (argc=<value optimized out>, argv=<value optimized out>,

envp=<value optimized out>) at fio.c:33

Apologize for late to reply.

Quoted 4 lines · Han Pingtian

Hey there,

When trying to run fio on one of our power system, segmentation fault

Can you give us the kind of cpu you are using ? (/proc/cpuinfo) I'm not used with PPC but maybe your processor doesn't support ATBU call on mfspr.

Erwan,