| Message ID | 54CA61DF.30807@kernel.dk |
|---|---|
| State | Not Applicable |
| Delegated to: | David Miller |
| Headers | show |
> > > > > > > > Yes, that does look like the case. Do you have a good trick on > > > > > > > > how > > > > > > > > to allocate a map for the highest possible cpu number without > > > > > > > > first > > > > > > > > iterating the cpu map? I couldn't find something that looks > > > > > > > > like a > > > > > > > > highest_possible_cpu() helper. > > > > > > > > > > > > > > Honestly I think that num_posible_cpus() should return the max of > > > > > > > number of CPUs (weigt), and the highest numbered CPU. It's a pain > > > > > > > in > > > > > > > the butt to handle this otherwise. > > > > > > > > > > > > Hear, hear!!! That would make my life easier, and would make this > > > > > > sort > > > > > > of problem much less likely to occur! > > > > > > > > > > How about this one? > > > > > > > > It make the machine work. > > > > > > Thanks for testing! > > > > > > > What's the status of this fix? It is still not applied on yesterdays > > 3.19.0-rc6-00105-gc59c961 git... > > Hmm, I thought commit a33c1ba29138 took care of it... Does the attached work? Sorry for taking so long - now I am back to the machine and turned it on yesterday. The machine is Sun Fire 3500 with 4 sparsely numbered 4 CPUs (6,7,18,19). First I fetched 4.2.0 and tried it without any patches. That got the following error - seems to be blk-mq related but I do not rmemeber if it is exactly the same as before. Next I tried 4.2 + your previously sent map-sz.patch but it makes no difference. spitfire_data_access_exception: SFSR[0000000000801009] SFAR[fffffdff01b02310], going. \|/ ____ \|/ "@'/ .. \`@" /_| \__/ |_\ \__U_/ swapper/6(1): Dax [#1] CPU: 19 PID: 1 Comm: swapper/6 Not tainted 4.2.0 #67 task: fffff800fe092fe0 ti: fffff800fe0b8000 task.ti: fffff800fe0b8000 TSTATE: 0000000080001602 TPC: 000000000065bad4 TNPC: 000000000065bad8 Y: 00000003 Not tainted TPC: <kobject_init+0x14/0xa0> g0: 0000000000000022 g1: 0000000080000000 g2: 0000000000000000 g3: 0000000000000001 g4: fffff800fe092fe0 g5: fffff800fe404000 g6: fffff800fe0b8000 g7: 0000000000000017 o0: 0000000000000000 o1: fffff800fe359bd0 o2: 00000000009c0c00 o3: 0000000000a2ba00 o4: fffff800fd50a368 o5: 0000000000000000 sp: fffff800fe0bafe1 ret_pc: 000000000065cd4c RPC: <kobject_uevent_env+0x4c/0x500> l0: fffff800fe092fe0 l1: fffff800fe042420 l2: 00000000008b15f0 l3: fffff800fe0bb9e0 l4: 00000000009bc415 l5: fffff8017e0bb9df l6: 0000000000000000 l7: fffff800fe193860 i0: 00000dff01b022d8 i1: 0000000000a4afe0 i2: 0000000000000000 i3: 0000000000a4aa28 i4: 00000000009c1448 i5: fffff800fe359bd0 i6: fffff800fe0bb091 i7: 000000000064a2c0 I7: <blk_mq_register_disk+0xe0/0x1a0> Call Trace: [000000000064a2c0] blk_mq_register_disk+0xe0/0x1a0 [000000000063f880] blk_register_queue+0xa0/0x120 [000000000064dbfc] add_disk+0x33c/0x480 [00000000006f3bd0] loop_add+0x190/0x280 [0000000000a8c5b0] loop_init+0x160/0x1b0 [0000000000426ea4] do_one_initcall+0xe4/0x1e0 [0000000000a70b8c] kernel_init_freeable+0x130/0x1e0 [000000000087cfa4] kernel_init+0x4/0x100 [0000000000406124] ret_from_fork+0x1c/0x2c [0000000000000000] (null) Caller[000000000064a2c0]: blk_mq_register_disk+0xe0/0x1a0 Caller[000000000063f880]: blk_register_queue+0xa0/0x120 Caller[000000000064dbfc]: add_disk+0x33c/0x480 Caller[00000000006f3bd0]: loop_add+0x190/0x280 Caller[0000000000a8c5b0]: loop_init+0x160/0x1b0 Caller[0000000000426ea4]: do_one_initcall+0xe4/0x1e0 Caller[0000000000a70b8c]: kernel_init_freeable+0x130/0x1e0 Caller[000000000087cfa4]: kernel_init+0x4/0x100 Caller[0000000000406124]: ret_from_fork+0x1c/0x2c Caller[0000000000000000]: (null) Instruction DUMP: 15002703 02c6401a 03200000 <c45e2038> 82088001 22c84009 c20e203c 11002704 92100 Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000009 Press Stop-A (L1-A) to return to the boot prom
diff --git a/block/blk-mq-cpumap.c b/block/blk-mq-cpumap.c index 5f13f4d0bcce..527d315dc1a5 100644 --- a/block/blk-mq-cpumap.c +++ b/block/blk-mq-cpumap.c @@ -88,10 +88,11 @@ int blk_mq_update_queue_map(unsigned int *map, unsigned int nr_queues) unsigned int *blk_mq_make_queue_map(struct blk_mq_tag_set *set) { unsigned int *map; + size_t sz; /* If cpus are offline, map them to first hctx */ - map = kzalloc_node(sizeof(*map) * nr_cpu_ids, GFP_KERNEL, - set->numa_node); + sz = max_t(unsigned int, nr_cpu_ids, num_possible_cpus()); + map = kzalloc_node(sizeof(*map) * sz, GFP_KERNEL, set->numa_node); if (!map) return NULL;