diff mbox series

[v2] powerpc/hv-gpci: Fix the H_GET_PERF_COUNTER_INFO hcall return value checks

Message ID 20240229122847.101162-1-kjain@linux.ibm.com (mailing list archive)
State Accepted
Commit ad86d7ee43b22aa2ed60fb982ae94b285c1be671
Headers show
Series [v2] powerpc/hv-gpci: Fix the H_GET_PERF_COUNTER_INFO hcall return value checks | expand

Checks

Context Check Description
snowpatch_ozlabs/github-powerpc_selftests success Successfully ran 8 jobs.
snowpatch_ozlabs/github-powerpc_ppctests success Successfully ran 8 jobs.
snowpatch_ozlabs/github-powerpc_sparse success Successfully ran 4 jobs.
snowpatch_ozlabs/github-powerpc_clang success Successfully ran 6 jobs.
snowpatch_ozlabs/github-powerpc_kernel_qemu success Successfully ran 23 jobs.

Commit Message

Kajol Jain Feb. 29, 2024, 12:28 p.m. UTC 
Running event hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/
in one of the system throws below error:

 ---Logs---
 # perf list | grep hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles
  hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=?/[Kernel PMU event]

 # perf stat -v -e hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/ sleep 2
Using CPUID 00800200
Control descriptor is not initialized
Warning:
hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/ event is not supported by the kernel.
failed to read counter hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/

 Performance counter stats for 'system wide':

   <not supported>      hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/

       2.000700771 seconds time elapsed

The above error is because of the hcall failure as required
permission "Enable Performance Information Collection" is not set.
Based on current code, single_gpci_request function did not check the
error type incase hcall fails and by default returns EINVAL. But we can
have other reasons for hcall failures like H_AUTHORITY/H_PARAMETER with
detail_rc as GEN_BUF_TOO_SMALL, for which we need to act accordingly.

Fix this issue by adding new checks in the single_gpci_request and
h_gpci_event_init functions.

Result after fix patch changes:

 # perf stat -e hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/ sleep 2
Error:
No permission to enable hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/ event.

Fixes: 220a0c609ad1 ("powerpc/perf: Add support for the hv gpci (get performance counter info) interface")
Reported-by: Akanksha J N <akanksha@linux.ibm.com>
Signed-off-by: Kajol Jain <kjain@linux.ibm.com>
---
Changelog:

v1 -> v2
- To make sure the hcall failure with H_PARAMETER is only because of
  buffer size issue, add check for detail_rc value to be
  GEN_BUF_TOO_SMALL(0x1B) as suggested by Michael Ellerman.

 arch/powerpc/perf/hv-gpci.c | 29 +++++++++++++++++++++++++++--
 1 file changed, 27 insertions(+), 2 deletions(-)

Comments

Michael Ellerman March 13, 2024, 1:19 p.m. UTC | #1 
On Thu, 29 Feb 2024 17:58:47 +0530, Kajol Jain wrote:
> Running event hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=0/
> in one of the system throws below error:
> 
>  ---Logs---
>  # perf list | grep hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles
>   hv_gpci/dispatch_timebase_by_processor_processor_time_in_timebase_cycles,phys_processor_idx=?/[Kernel PMU event]
> 
> [...]

Applied to powerpc/next.

[1/1] powerpc/hv-gpci: Fix the H_GET_PERF_COUNTER_INFO hcall return value checks
      https://git.kernel.org/powerpc/c/ad86d7ee43b22aa2ed60fb982ae94b285c1be671

cheers
diff mbox series

Patch

diff --git a/arch/powerpc/perf/hv-gpci.c b/arch/powerpc/perf/hv-gpci.c
index 27f18119fda1..89bfdc2ce8bc 100644
--- a/arch/powerpc/perf/hv-gpci.c
+++ b/arch/powerpc/perf/hv-gpci.c
@@ -695,6 +695,20 @@  static unsigned long single_gpci_request(u32 req, u32 starting_index,
 
 	ret = plpar_hcall_norets(H_GET_PERF_COUNTER_INFO,
 			virt_to_phys(arg), HGPCI_REQ_BUFFER_SIZE);
+
+	/*
+	 * ret value as 'H_PARAMETER' with detail_rc as 'GEN_BUF_TOO_SMALL',
+	 * specifies that the current buffer size cannot accommodate
+	 * all the information and a partial buffer returned.
+	 * Since in this function we are only accessing data for a given starting index,
+	 * we don't need to accommodate whole data and can get required count by
+	 * accessing first entry data.
+	 * Hence hcall fails only incase the ret value is other than H_SUCCESS or
+	 * H_PARAMETER with detail_rc value as GEN_BUF_TOO_SMALL(0x1B).
+	 */
+	if (ret == H_PARAMETER && be32_to_cpu(arg->params.detail_rc) == 0x1B)
+		ret = 0;
+
 	if (ret) {
 		pr_devel("hcall failed: 0x%lx\n", ret);
 		goto out;
@@ -759,6 +773,7 @@  static int h_gpci_event_init(struct perf_event *event)
 {
 	u64 count;
 	u8 length;
+	unsigned long ret;
 
 	/* Not our event */
 	if (event->attr.type != event->pmu->type)
@@ -789,13 +804,23 @@  static int h_gpci_event_init(struct perf_event *event)
 	}
 
 	/* check if the request works... */
-	if (single_gpci_request(event_get_request(event),
+	ret = single_gpci_request(event_get_request(event),
 				event_get_starting_index(event),
 				event_get_secondary_index(event),
 				event_get_counter_info_version(event),
 				event_get_offset(event),
 				length,
-				&count)) {
+				&count);
+
+	/*
+	 * ret value as H_AUTHORITY implies that partition is not permitted to retrieve
+	 * performance information, and required to set
+	 * "Enable Performance Information Collection" option.
+	 */
+	if (ret == H_AUTHORITY)
+		return -EPERM;
+
+	if (ret) {
 		pr_devel("gpci hcall failed\n");
 		return -EINVAL;
 	}