Skip to content

Commit 83cc588

Browse files
Suzuki K Poulosemathieupoirier
Suzuki K Poulose
authored andcommitted
coresight: tmc-etr: alloc_perf_buf: Do not call smp_processor_id from preemptible
During a perf session we try to allocate buffers on the "node" associated with the CPU the event is bound to. If it is not bound to a CPU, we use the current CPU node, using smp_processor_id(). However this is unsafe in a pre-emptible context and could generate the splats as below : BUG: using smp_processor_id() in preemptible [00000000] code: perf/1743 caller is tmc_alloc_etr_buffer+0x1bc/0x1f0 CPU: 1 PID: 1743 Comm: perf Not tainted 5.1.0-rc6-147786-g116841e torvalds#344 Hardware name: ARM LTD ARM Juno Development Platform/ARM Juno Development Platform, BIOS EDK II Feb 1 2019 Call trace: dump_backtrace+0x0/0x150 show_stack+0x14/0x20 dump_stack+0x9c/0xc4 debug_smp_processor_id+0x10c/0x110 tmc_alloc_etr_buffer+0x1bc/0x1f0 etm_setup_aux+0x1c4/0x230 rb_alloc_aux+0x1b8/0x2b8 perf_mmap+0x35c/0x478 mmap_region+0x34c/0x4f0 do_mmap+0x2d8/0x418 vm_mmap_pgoff+0xd0/0xf8 ksys_mmap_pgoff+0x88/0xf8 __arm64_sys_mmap+0x28/0x38 el0_svc_handler+0xd8/0x138 el0_svc+0x8/0xc Use NUMA_NO_NODE hint instead of using the current node for events not bound to CPUs. Fixes: 22f429f ("coresight: etm-perf: Add support for ETR backend") Cc: Mathieu Poirier <[email protected]> Signed-off-by: Suzuki K Poulose <[email protected]> Signed-off-by: Mathieu Poirier <[email protected]>
1 parent 2b3fa03 commit 83cc588

File tree

1 file changed

+2
-5
lines changed

1 file changed

+2
-5
lines changed

drivers/hwtracing/coresight/coresight-tmc-etr.c

+2-5
Original file line numberDiff line numberDiff line change
@@ -1184,14 +1184,11 @@ static struct etr_buf *
11841184
alloc_etr_buf(struct tmc_drvdata *drvdata, struct perf_event *event,
11851185
int nr_pages, void **pages, bool snapshot)
11861186
{
1187-
int node, cpu = event->cpu;
1187+
int node;
11881188
struct etr_buf *etr_buf;
11891189
unsigned long size;
11901190

1191-
if (cpu == -1)
1192-
cpu = smp_processor_id();
1193-
node = cpu_to_node(cpu);
1194-
1191+
node = (event->cpu == -1) ? NUMA_NO_NODE : cpu_to_node(event->cpu);
11951192
/*
11961193
* Try to match the perf ring buffer size if it is larger
11971194
* than the size requested via sysfs.

0 commit comments

Comments
 (0)