Messages in this thread | | | Date | Sun, 22 Jun 2008 22:58:05 +0200 | From | "Vegard Nossum" <> | Subject | Re: [PATCH -tip-rcu] Make rcutorture more vicious: make quiescent rcutorture less power-hungry |
| |
On Sun, Jun 22, 2008 at 10:06 PM, Paul E. McKenney <paulmck@linux.vnet.ibm.com> wrote: > This patch makes the non-module rcutorture a bit more friendly to > the power-conservation code. This is a rather simple-minded approach. > More sophisticated approaches would get rid of the rcutorture tasks > while rcutorture execution was suppressed, but attempts thus far to > do this have not gone well -- calling rcu_torture_init() from a /proc > callout results in oopses.
Hi Paul,
I applied your three patches
[PATCH] Make rcutorture more vicious: add stutter feature [PATCH] Make rcutorture more vicious: reinstate boot-time testing [PATCH -tip-rcu] Make rcutorture more vicious: make quiescent rcutorture less power-hungry
to v2.6.26-rc7 and gave it a quick testing in qemu. But it seems to hang during gdb self-tests at boot:
rcu-torture:--- Start of test: nreaders=6 nfakewriters=4 stat_interval=0 verbose=0 test_no_idle_hz=0 shuffle_interval=3 stutter=5 Clocksource tsc unstable (delta = 617797275 ns) Total HugeTLB memory allocated, 0 Installing knfsd (copyright (C) 1996 okir@monad.swb.de). NTFS driver 2.1.29 [Flags: R/O]. msgmni has been set to 228 io scheduler noop registered io scheduler anticipatory registered io scheduler deadline registered io scheduler cfq registered (default) pci 0000:00:00.0: Limiting direct PCI/PCI transfers pci 0000:00:01.0: PIIX3: Enabling Passive Release pci 0000:00:01.0: Activating ISA DMA hang workarounds pci 0000:00:02.0: Boot video device Real Time Clock Driver v1.12ac Linux agpgart interface v0.103 Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing disabled serial8250: ttyS0 at I/O 0x3f8 (irq = 4) is a 16450 FDC 0 is a S82078B brd: module loaded loop: module loaded kgdb: Registered I/O driver kgdbts. kgdbts:RUN plant and detach test
<--- full stop
gdb shows this:
kgdb_handle_exception (evector=0, signo=<value optimized out>, ecode=<value optimized out>, regs=0xc0769eec) at include/asm/processor.h:688 688 asm volatile("rep; nop" ::: "memory"); (gdb) bt #0 kgdb_handle_exception (evector=0, signo=<value optimized out>, ecode=<value optimized out>, regs=0xc0769eec) at include/asm/processor.h:688 #1 0xc011ffa1 in kgdb_notify (self=<value optimized out>, cmd=9, ptr=0xc0769e30) at arch/x86/kernel/kgdb.c:478 #2 0xc014f527 in notifier_call_chain (nl=<value optimized out>, val=9, v=0xc0769e30, nr_to_call=-1, nr_calls=0x0) at kernel/notifier.c:85 #3 0xc014f925 in __atomic_notifier_call_chain (nh=<value optimized out>, val=9, v=0xc0769e30, nr_to_call=-1, nr_calls=0x0) at kernel/notifier.c:174 #4 0xc014f95a in atomic_notifier_call_chain (nh=0x2, val=0, v=0xffffffff) at kernel/notifier.c:183 #5 0xc014f98d in notify_die (val=0, str=<value optimized out>, regs=<value optimized out>, err=0, trap=0, sig=8) at kernel/notifier.c:564 #6 0xc010a8e8 in do_divide_error (regs=0xc0769eec, error_code=0) at arch/x86/kernel/traps_32.c:594 #7 0xc058cdba in page_fault () #8 0xc080f780 in per_cpu__init_tss () #9 0xc013c255 in do_softirq () at kernel/softirq.c:271 #10 0xc013c255 in do_softirq () at kernel/softirq.c:271 #11 0xc013c3f5 in irq_exit () at kernel/softirq.c:310 #12 0xc011db28 in smp_apic_timer_interrupt (regs=<value optimized out>) at arch/x86/kernel/apic_32.c:619 #13 0xc0108ed3 in apic_timer_interrupt () at include/asm/desc.h:309 #14 0xc0106ce0 in ?? () at arch/x86/kernel/process_32.c:466 #15 0xc0768000 in ?? () #16 0x00000000 in ?? ()
Without your patches, the kernel gets through here just fine.
These are my RCU-related options:
$ grep RCU .config # CONFIG_CLASSIC_RCU is not set CONFIG_PREEMPT_RCU=y CONFIG_RCU_TRACE=y CONFIG_RCU_TORTURE_TEST=y CONFIG_RCU_TORTURE_TEST_RUNNABLE=y
As a convenience feature, it would be nice to have a boot option controlling the run-at-boot as well, simply because it saves a recompile for me :-) Actually, for some reason, the test still starts at boot even when CONFIG_RCU_TORTURE_TEST_RUNNABLE=n. Is a hunk missing from patch #2?
Vegard
-- "The animistic metaphor of the bug that maliciously sneaked in while the programmer was not looking is intellectually dishonest as it disguises that the error is the programmer's own creation." -- E. W. Dijkstra, EWD1036
| |