Patch "perf arm-spe: Use old behavior when opening old SPE files" has been added to the 6.12-stable tree

[gregkh at linuxfoundation.org]

This is a note to let you know that I've just added the patch titled

    perf arm-spe: Use old behavior when opening old SPE files

to the 6.12-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     perf-arm-spe-use-old-behavior-when-opening-old-spe-files.patch
and it can be found in the queue-6.12 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable at vger.kernel.org> know about it.


>From ba993e5ada1ddce7a71140dc85ef65bc2cd981bc Mon Sep 17 00:00:00 2001
From: James Clark <james.clark at linaro.org>
Date: Tue, 29 Oct 2024 14:37:33 +0000
Subject: perf arm-spe: Use old behavior when opening old SPE files

From: James Clark <james.clark at linaro.org>

commit ba993e5ada1ddce7a71140dc85ef65bc2cd981bc upstream.

Since the linked commit, we stopped interpreting data source if the
perf.data file doesn't have the new metadata version. This means that
perf c2c will show no samples in this case.

Keep the old behavior so old files can be opened, but also still show
the new warning that updating might improve the decoding.

Also re-write the warning to be more concise and specific to a user.

Fixes: ba5e7169e548 ("perf arm-spe: Use metadata to decide the data source feature")
Signed-off-by: James Clark <james.clark at linaro.org>
Reviewed-by: Leo Yan <leo.yan at arm.com>
Cc: Julio.Suarez at arm.com
Cc: Kiel.Friedt at arm.com
Cc: Ryan.Roberts at arm.com
Cc: Will Deacon <will at kernel.org>
Cc: Mike Leach <mike.leach at linaro.org>
Cc: linux-arm-kernel at lists.infradead.org
Cc: Besar Wicaksono <bwicaksono at nvidia.com>
Cc: John Garry <john.g.garry at oracle.com>
Link: https://lore.kernel.org/r/20241029143734.291638-1-james.clark@linaro.org
Signed-off-by: Namhyung Kim <namhyung at kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh at linuxfoundation.org>
---
 tools/perf/util/arm-spe.c |   50 ++++++++++++++++++++++------------------------
 1 file changed, 24 insertions(+), 26 deletions(-)

--- a/tools/perf/util/arm-spe.c
+++ b/tools/perf/util/arm-spe.c
@@ -655,37 +655,35 @@ static bool arm_spe__is_common_ds_encodi
 	u64 *metadata = NULL;
 	u64 midr = 0;
 
-	/*
-	 * Metadata version 1 doesn't contain any info for MIDR.
-	 * Simply return false in this case.
-	 */
+	/* Metadata version 1 assumes all CPUs are the same (old behavior) */
 	if (spe->metadata_ver == 1) {
-		pr_warning_once("The data file contains metadata version 1, "
-				"which is absent the info for data source. "
-				"Please upgrade the tool to record data.\n");
-		return false;
-	}
-
-	/* CPU ID is -1 for per-thread mode */
-	if (speq->cpu < 0) {
-		/*
-		 * On the heterogeneous system, due to CPU ID is -1,
-		 * cannot confirm the data source packet is supported.
-		 */
-		if (!spe->is_homogeneous)
-			return false;
+		const char *cpuid;
 
-		/* In homogeneous system, simply use CPU0's metadata */
-		if (spe->metadata)
-			metadata = spe->metadata[0];
+		pr_warning_once("Old SPE metadata, re-record to improve decode accuracy\n");
+		cpuid = perf_env__cpuid(spe->session->evlist->env);
+		midr = strtol(cpuid, NULL, 16);
 	} else {
-		metadata = arm_spe__get_metadata_by_cpu(spe, speq->cpu);
-	}
+		/* CPU ID is -1 for per-thread mode */
+		if (speq->cpu < 0) {
+			/*
+			 * On the heterogeneous system, due to CPU ID is -1,
+			 * cannot confirm the data source packet is supported.
+			 */
+			if (!spe->is_homogeneous)
+				return false;
+
+			/* In homogeneous system, simply use CPU0's metadata */
+			if (spe->metadata)
+				metadata = spe->metadata[0];
+		} else {
+			metadata = arm_spe__get_metadata_by_cpu(spe, speq->cpu);
+		}
 
-	if (!metadata)
-		return false;
+		if (!metadata)
+			return false;
 
-	midr = metadata[ARM_SPE_CPU_MIDR];
+		midr = metadata[ARM_SPE_CPU_MIDR];
+	}
 
 	is_in_cpu_list = is_midr_in_range_list(midr, common_ds_encoding_cpus);
 	if (is_in_cpu_list)


Patches currently in stable-queue which might be from james.clark at linaro.org are

queue-6.12/perf-arm-spe-use-arm_spe_op_branch_eret-when-synthes.patch
queue-6.12/perf-arm-spe-correctly-set-sample-flags.patch
queue-6.12/perf-cs-etm-queue-context-packets-for-frontend.patch
queue-6.12/perf-arm-spe-add-branch-stack.patch
queue-6.12/perf-arm-spe-use-metadata-to-decide-the-data-source-.patch
queue-6.12/perf-pmu-skip-test-on-arm64-when-slots-is-zero.patch
queue-6.12/perf-sched-add-missing-mmap2-handler-in-timehist.patch
queue-6.12/perf-synthetic-events-fix-stale-build-id-in-module-mmap2-records.patch
queue-6.12/perf-inject-add-convert-callchain-option.patch
queue-6.12/perf-build-id-ensure-snprintf-string-is-empty-when-size-is-0.patch
queue-6.12/perf-debuginfo-fix-libdw-api-contract-violations.patch
queue-6.12/coresight-etm4x-correct-trcvmidcctlr1-save-and-resto.patch
queue-6.12/perf-event-fix-size-of-synthesized-sample-with-branc.patch
queue-6.12/perf-remove-redundant-kernel.h-include.patch
queue-6.12/perf-fix-off-by-one-stack-buffer-overflow-in-kallsym.patch
queue-6.12/perf-inject-with-convert-callchain-ignore-the-dummy-event-for-dwarf-stacks.patch
queue-6.12/perf-maps-add-maps__mutate_mapping.patch
queue-6.12/perf-arm-spe-use-old-behavior-when-opening-old-spe-files.patch