[PATCH 6.1.y v2 00/18] Backport fixes for -Wdiscarded-qualifiers and -Wnonnull with newer glibc

[PATCH 6.1.y v2 00/18] Backport fixes for -Wdiscarded-qualifiers and -Wnonnull with newer glibc

[Shreenidhi Shedi]

Hi all,

This patch series backports a number of patches from master to 6.1.y
to fix `-Wdiscarded-qualifiers` and `-Wnonnull` build issues with
newer glibc versions.

I will port these changes to other stable trees once this gets reviewed.

Arnaldo Carvalho de Melo (12):
  perf diff: Constify strchr() return variables
  perf test bpf: Address error about non-null argument for epoll_pwait
    2nd arg
  perf trace: Deal with compiler const checks
  perf tools: Use const for variables receiving str{str,r?chr}() returns
  perf strlist: Don't write to const memory
  perf metricgroup: Constify variables storing the result of strchr() on
    const tables
  perf session: Don't write to memory pointed to a const pointer
  perf trace-event: Constify variables storing the result of strchr() on
    const tables
  perf units: Constify variables storing the result of strchr() on const
    tables
  perf time-utils: Constify variables storing the result of strchr() on
    const tables
  perf demangle-java: Constify variables storing the result of strchr()
    on const tables
  perf bpf-event: Constify variables storing the result of strchr() on
    const tables

Dr. David Alan Gilbert (1):
  perf tools: Remove unused color_fwrite_lines

Mikhail Gavrilov (1):
  libbpf: Fix -Wdiscarded-qualifiers under C23

The following patches by me are branch specific. Upstream code is
refactored and the troubling piece of code is not there anymore.

Shreenidhi Shedi (4):
  perf list: Fix -Wdiscarded-qualifiers under C23
  perf parse-events: Fix -Wdiscarded-qualifiers under C23
  perf bpf: Fix -Wdiscarded-qualifiers under C23
  perf parse-events:: Fix -Wdiscarded-qualifiers under C23

 tools/lib/bpf/libbpf.c             |  2 +-
 tools/perf/builtin-diff.c          | 12 +++++-------
 tools/perf/builtin-list.c          |  3 ++-
 tools/perf/builtin-trace.c         |  2 +-
 tools/perf/jvmti/libjvmti.c        |  2 +-
 tools/perf/tests/bpf.c             |  3 ++-
 tools/perf/util/bpf-event.c        |  3 ++-
 tools/perf/util/bpf-loader.c       |  2 +-
 tools/perf/util/color.c            | 28 ----------------------------
 tools/perf/util/color.h            |  1 -
 tools/perf/util/demangle-java.c    |  2 +-
 tools/perf/util/evlist.c           |  3 ++-
 tools/perf/util/metricgroup.c      |  3 +--
 tools/perf/util/parse-events.c     |  2 +-
 tools/perf/util/print-events.c     |  4 ++--
 tools/perf/util/session.c          |  6 +++---
 tools/perf/util/strlist.c          | 12 ++++++++----
 tools/perf/util/time-utils.c       |  4 ++--
 tools/perf/util/trace-event-info.c |  2 +-
 tools/perf/util/units.c            |  2 +-
 20 files changed, 37 insertions(+), 61 deletions(-)

v1:
https://lore.kernel.org/all/20260509173559.10999-1-yesshedi@gmail.com/

v2:
Address comment from Greg
https://lore.kernel.org/all/2026051001-jacket-reliance-3ec6@gregkh/

-- 
2.54.0

[PATCH 6.1.y v2 01/18] libbpf: Fix -Wdiscarded-qualifiers under C23

[Shreenidhi Shedi]

From: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>

commit d70f79fef65810faf64dbae1f3a1b5623cdb2345 upstream

glibc ≥ 2.42 (GCC 15) defaults to -std=gnu23, which promotes
-Wdiscarded-qualifiers to an error.

In C23, strstr() and strchr() return "const char *".

Change variable types to const char * where the pointers are never
modified (res, sym_sfx, next_path).

Suggested-by: Florian Weimer <fweimer@redhat.com>
Suggested-by: Andrii Nakryiko <andrii@kernel.org>
Signed-off-by: Mikhail Gavrilov <mikhail.v.gavrilov@gmail.com>
Link: https://lore.kernel.org/r/20251206092825.1471385-1-mikhail.v.gavrilov@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/lib/bpf/libbpf.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
index 7bd6aff6e260..33b214a91338 100644
--- a/tools/lib/bpf/libbpf.c
+++ b/tools/lib/bpf/libbpf.c
@@ -10748,7 +10748,7 @@ static int resolve_full_path(const char *file, char *result, size_t result_sz)
 		if (!search_paths[i])
 			continue;
 		for (s = search_paths[i]; s != NULL; s = strchr(s, ':')) {
-			char *next_path;
+			const char *next_path;
 			int seg_len;
 
 			if (s[0] == ':')
-- 
2.54.0

[PATCH 6.1.y v2 02/18] perf diff: Constify strchr() return variables

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@kernel.org>

commit f6f41aef53761517391b6192fe5b4bc30b2d717a upstream

Newer glibc versions return const char for strchr() when the 's' arg is
const, change the return variable to const to match that.

Also we don't need to turn that ',' into a '\0', as strtol() will stop
in the first invalid char. No need to touch read only memory.

First noticed with fedora 44.

Reviewed-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20251211221756.96294-3-acme@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/builtin-diff.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/tools/perf/builtin-diff.c b/tools/perf/builtin-diff.c
index ed07cc6cca56..9a05d67f541b 100644
--- a/tools/perf/builtin-diff.c
+++ b/tools/perf/builtin-diff.c
@@ -177,10 +177,9 @@ static struct header_column {
 	}
 };
 
-static int setup_compute_opt_wdiff(char *opt)
+static int setup_compute_opt_wdiff(const char *opt)
 {
-	char *w1_str = opt;
-	char *w2_str;
+	const char *w1_str = opt, *w2_str;
 
 	int ret = -EINVAL;
 
@@ -191,8 +190,7 @@ static int setup_compute_opt_wdiff(char *opt)
 	if (!w2_str)
 		goto out;
 
-	*w2_str++ = 0x0;
-	if (!*w2_str)
+	if (!*++w2_str)
 		goto out;
 
 	compute_wdiff_w1 = strtol(w1_str, NULL, 10);
@@ -213,7 +211,7 @@ static int setup_compute_opt_wdiff(char *opt)
 	return ret;
 }
 
-static int setup_compute_opt(char *opt)
+static int setup_compute_opt(const char *opt)
 {
 	if (compute == COMPUTE_WEIGHTED_DIFF)
 		return setup_compute_opt_wdiff(opt);
@@ -233,7 +231,7 @@ static int setup_compute(const struct option *opt, const char *str,
 	char *cstr = (char *) str;
 	char buf[50];
 	unsigned i;
-	char *option;
+	const char *option;
 
 	if (!str) {
 		*cp = COMPUTE_DELTA;
-- 
2.54.0

[PATCH 6.1.y v2 03/18] perf test bpf: Address error about non-null argument for epoll_pwait 2nd arg

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit ed847e30f001b207013b6136c264454d7560557f upstream

First noticed on Fedora Rawhide:

  tests/bpf.c: In function ‘epoll_pwait_loop’:
  tests/bpf.c:36:17: error: argument 2 null where non-null expected [-Werror=nonnull]
     36 |                 epoll_pwait(-(i + 1), NULL, 0, 0, NULL);
        |                 ^~~~~~~~~~~
  In file included from tests/bpf.c:5:
  /usr/include/sys/epoll.h:134:12: note: in a call to function ‘epoll_pwait’ declared ‘nonnull’
    134 | extern int epoll_pwait (int __epfd, struct epoll_event *__events,
        |            ^~~~~~~~~~~

  [perfbuilder@27cfe44d67ed perf-6.5.0-rc2]$ gcc -v
  Using built-in specs.
  COLLECT_GCC=gcc
  COLLECT_LTO_WRAPPER=/usr/libexec/gcc/x86_64-redhat-linux/13/lto-wrapper
  OFFLOAD_TARGET_NAMES=nvptx-none
  OFFLOAD_TARGET_DEFAULT=1
  Target: x86_64-redhat-linux
  Configured with: ../configure --enable-bootstrap --enable-languages=c,c++,fortran,objc,obj-c++,ada,go,d,m2,lto --prefix=/usr --mandir=/usr/share/man --infodir=/usr/share/info --with-bugurl=http://bugzilla.redhat.com/bugzilla --enable-shared --enable-threads=posix --enable-checking=release --enable-multilib --with-system-zlib --enable-__cxa_atexit --disable-libunwind-exceptions --enable-gnu-unique-object --enable-linker-build-id --with-gcc-major-version-only --enable-libstdcxx-backtrace --with-libstdcxx-zoneinfo=/usr/share/zoneinfo --with-linker-hash-style=gnu --enable-plugin --enable-initfini-array --with-isl=/builddir/build/BUILD/gcc-13.2.1-20230728/obj-x86_64-redhat-linux/isl-install --enable-offload-targets=nvptx-none --without-cuda-driver --enable-offload-defaulted --enable-gnu-indirect-function --enable-cet --with-tune=generic --with-arch_32=i686 --build=x86_64-redhat-linux --with-build-config=bootstrap-lto --enable-link-serialization=1
  Thread model: posix
  Supported LTO compression algorithms: zlib zstd
  gcc version 13.2.1 20230728 (Red Hat 13.2.1-1) (GCC)
  [perfbuilder@27cfe44d67ed perf-6.5.0-rc2]$

Just add that argument to address this compiler warning.

Cc: Adrian Hunter <adrian.hunter@intel.com>
Cc: Ian Rogers <irogers@google.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Link: https://lore.kernel.org/lkml/ZMj8+bvN86D0ZKiB@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/tests/bpf.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/perf/tests/bpf.c b/tools/perf/tests/bpf.c
index 6a4235a9cf57..d5bf0e47a48a 100644
--- a/tools/perf/tests/bpf.c
+++ b/tools/perf/tests/bpf.c
@@ -29,11 +29,12 @@
 
 static int epoll_pwait_loop(void)
 {
+	struct epoll_event events;
 	int i;
 
 	/* Should fail NR_ITERS times */
 	for (i = 0; i < NR_ITERS; i++)
-		epoll_pwait(-(i + 1), NULL, 0, 0, NULL);
+		epoll_pwait(-(i + 1), &events, 0, 0, NULL);
 	return 0;
 }
 
-- 
2.54.0

[PATCH 6.1.y v2 04/18] perf list: Fix -Wdiscarded-qualifiers under C23

[Shreenidhi Shedi]

builtin-list.c:104:31: warning: assignment discards 'const' qualifier
from pointer target type [-Wdiscarded-qualifiers]
  104 |                 else if ((sep = strchr(argv[i], ':')) != NULL) {

glibc >= 2.42 defaults to -std=gnu23, which promotes
    -Wdiscarded-qualifiers to an error.

Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/builtin-list.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-list.c b/tools/perf/builtin-list.c
index 58e1ec1654ef..a37f3b8a597c 100644
--- a/tools/perf/builtin-list.c
+++ b/tools/perf/builtin-list.c
@@ -75,7 +75,8 @@ int cmd_list(int argc, const char **argv)
 	}
 
 	for (i = 0; i < argc; ++i) {
-		char *sep, *s;
+		const char *sep;
+		char *s;
 
 		if (strcmp(argv[i], "tracepoint") == 0)
 			print_tracepoint_events(NULL, NULL, raw_dump);
-- 
2.54.0

[PATCH 6.1.y v2 05/18] perf trace: Deal with compiler const checks

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit 2c850606a46b319d5128bda59f67b1fc642d94ef upstream

The strchr() function these days return const/non-const based on the arg
it receives, and sometimes we need to use casts when we're dealing with
variables that are used in code that needs to safely change the returned
value and sometimes not (as it points to really const areas).

Tweak one such case.

Reviewed-by: Ian Rogers <irogers@google.com>
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/builtin-trace.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c
index 441655e659c2..313eb0929324 100644
--- a/tools/perf/builtin-trace.c
+++ b/tools/perf/builtin-trace.c
@@ -4622,7 +4622,7 @@ static int trace__parse_events_option(const struct option *opt, const char *str,
 	}
 
 	while (1) {
-		if ((sep = strchr(s, ',')) != NULL)
+		if ((sep = strchr((char *)s, ',')) != NULL)
 			*sep = '\0';
 
 		list = 0;
-- 
2.54.0

[PATCH 6.1.y v2 06/18] perf tools: Use const for variables receiving str{str,r?chr}() returns

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@kernel.org>

commit 45718bce7daf39c618188b70a52644bb5a2f968a upstream

Newer glibc versions return const char for str{str,chr}() where the
haystack/s is const so to avoid warnings like these on fedora 44 change
some variables to const:

  36     8.17 fedora:44                     : FAIL gcc version 15.2.1 20251111 (Red Hat 15.2.1-4) (GCC)
    libbpf.c: In function 'kallsyms_cb':
    libbpf.c:8489:13: error: assignment discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers]
     8489 |         res = strstr(sym_name, ".llvm.");

Reviewed-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20251211221756.96294-4-acme@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/jvmti/libjvmti.c | 2 +-
 tools/perf/util/evlist.c    | 3 ++-
 2 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/tools/perf/jvmti/libjvmti.c b/tools/perf/jvmti/libjvmti.c
index fcca275e5bf9..d6b04d0fd35a 100644
--- a/tools/perf/jvmti/libjvmti.c
+++ b/tools/perf/jvmti/libjvmti.c
@@ -142,7 +142,7 @@ copy_class_filename(const char * class_sign, const char * file_name, char * resu
 	*/
 	if (*class_sign == 'L') {
 		int j, i = 0;
-		char *p = strrchr(class_sign, '/');
+		const char *p = strrchr(class_sign, '/');
 		if (p) {
 			/* drop the 'L' prefix and copy up to the final '/' */
 			for (i = 0; i < (p - class_sign); i++)
diff --git a/tools/perf/util/evlist.c b/tools/perf/util/evlist.c
index 41bbe6f85b0d..391a694ae2af 100644
--- a/tools/perf/util/evlist.c
+++ b/tools/perf/util/evlist.c
@@ -1885,7 +1885,8 @@ static int evlist__parse_control_fifo(const char *str, int *ctl_fd, int *ctl_fd_
 
 int evlist__parse_control(const char *str, int *ctl_fd, int *ctl_fd_ack, bool *ctl_fd_close)
 {
-	char *comma = NULL, *endptr = NULL;
+	const char *comma = NULL;
+	char *endptr = NULL;
 
 	*ctl_fd_close = false;
 
-- 
2.54.0

[PATCH 6.1.y v2 07/18] perf parse-events: Fix -Wdiscarded-qualifiers under C23

[Shreenidhi Shedi]

glibc >= 2.42 defaults to -std=gnu23, which promotes
    -Wdiscarded-qualifiers to an error.

util/print-events.c:206:21: warning: assignment discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
  206 |                 buf = strchr(nd->s, '@');

util/print-events.c:215:29: warning: assignment discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]
  215 |                         ptr = strchr(nd2->s, '@');

Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/print-events.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/print-events.c b/tools/perf/util/print-events.c
index c4d5d87fae2f..1ef1a3ed8051 100644
--- a/tools/perf/util/print-events.c
+++ b/tools/perf/util/print-events.c
@@ -203,7 +203,7 @@ void print_sdt_events(const char *subsys_glob, const char *event_glob,
 	strlist__delete(bidlist);
 
 	strlist__for_each_entry(nd, sdtlist) {
-		buf = strchr(nd->s, '@');
+		buf = strchr((char *)nd->s, '@');
 		if (buf)
 			*(buf++) = '\0';
 		if (name_only) {
@@ -212,7 +212,7 @@ void print_sdt_events(const char *subsys_glob, const char *event_glob,
 		}
 		nd2 = strlist__next(nd);
 		if (nd2) {
-			ptr = strchr(nd2->s, '@');
+			ptr = strchr((char *)nd2->s, '@');
 			if (ptr)
 				*ptr = '\0';
 			if (strcmp(nd->s, nd2->s) == 0)
-- 
2.54.0

[PATCH 6.1.y v2 08/18] perf tools: Remove unused color_fwrite_lines

[Shreenidhi Shedi]

From: "Dr. David Alan Gilbert" <linux@treblig.org>

commit c7c1bb78f3eec716bc35f58d74592331cc3281b2 upstream

color_fwrite_lines() was added by 2009's commit
8fc0321f1ad0 ("perf_counter tools: Add color terminal output support")

but has never been used.

Remove it.

Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Reviewed-by: Ian Rogers <irogers@google.com>
Link: https://lore.kernel.org/r/20241009003938.254936-1-linux@treblig.org
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/color.c | 28 ----------------------------
 tools/perf/util/color.h |  1 -
 2 files changed, 29 deletions(-)

diff --git a/tools/perf/util/color.c b/tools/perf/util/color.c
index bffbdd216a6a..e51f0a676a22 100644
--- a/tools/perf/util/color.c
+++ b/tools/perf/util/color.c
@@ -93,34 +93,6 @@ int color_fprintf(FILE *fp, const char *color, const char *fmt, ...)
 	return r;
 }
 
-/*
- * This function splits the buffer by newlines and colors the lines individually.
- *
- * Returns 0 on success.
- */
-int color_fwrite_lines(FILE *fp, const char *color,
-		size_t count, const char *buf)
-{
-	if (!*color)
-		return fwrite(buf, count, 1, fp) != 1;
-
-	while (count) {
-		char *p = memchr(buf, '\n', count);
-
-		if (p != buf && (fputs(color, fp) < 0 ||
-				fwrite(buf, p ? (size_t)(p - buf) : count, 1, fp) != 1 ||
-				fputs(PERF_COLOR_RESET, fp) < 0))
-			return -1;
-		if (!p)
-			return 0;
-		if (fputc('\n', fp) < 0)
-			return -1;
-		count -= p + 1 - buf;
-		buf = p + 1;
-	}
-	return 0;
-}
-
 const char *get_percent_color(double percent)
 {
 	const char *color = PERF_COLOR_NORMAL;
diff --git a/tools/perf/util/color.h b/tools/perf/util/color.h
index 01f7bed21c9b..aecf56dae73f 100644
--- a/tools/perf/util/color.h
+++ b/tools/perf/util/color.h
@@ -39,7 +39,6 @@ int color_vsnprintf(char *bf, size_t size, const char *color,
 int color_vfprintf(FILE *fp, const char *color, const char *fmt, va_list args);
 int color_fprintf(FILE *fp, const char *color, const char *fmt, ...);
 int color_snprintf(char *bf, size_t size, const char *color, const char *fmt, ...);
-int color_fwrite_lines(FILE *fp, const char *color, size_t count, const char *buf);
 int value_color_snprintf(char *bf, size_t size, const char *fmt, double value);
 int percent_color_snprintf(char *bf, size_t size, const char *fmt, ...);
 int percent_color_len_snprintf(char *bf, size_t size, const char *fmt, ...);
-- 
2.54.0

[PATCH 6.1.y v2 09/18] perf strlist: Don't write to const memory

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit 678ed6b707e4b2db250f255d2f959322896dae65 upstream

Do a strdup to the list string and parse from it, free at the end.

This is to deal with newer glibcs const-correctness.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/strlist.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/tools/perf/util/strlist.c b/tools/perf/util/strlist.c
index 8a868cbeffae..98883672fcf4 100644
--- a/tools/perf/util/strlist.c
+++ b/tools/perf/util/strlist.c
@@ -139,21 +139,25 @@ static int strlist__parse_list_entry(struct strlist *slist, const char *s,
 	return err;
 }
 
-static int strlist__parse_list(struct strlist *slist, const char *s, const char *subst_dir)
+static int strlist__parse_list(struct strlist *slist, const char *list, const char *subst_dir)
 {
-	char *sep;
+	char *sep, *s = strdup(list), *sdup = s;
 	int err;
 
+	if (s == NULL)
+		return -ENOMEM;
+
 	while ((sep = strchr(s, ',')) != NULL) {
 		*sep = '\0';
 		err = strlist__parse_list_entry(slist, s, subst_dir);
-		*sep = ',';
 		if (err != 0)
 			return err;
 		s = sep + 1;
 	}
 
-	return *s ? strlist__parse_list_entry(slist, s, subst_dir) : 0;
+	err = *s ? strlist__parse_list_entry(slist, s, subst_dir) : 0;
+	free(sdup);
+	return err;
 }
 
 struct strlist *strlist__new(const char *list, const struct strlist_config *config)
-- 
2.54.0

[PATCH 6.1.y v2 10/18] perf metricgroup: Constify variables storing the result of strchr() on const tables

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit b42868624c7d00206f77d19a6fbfea73a44ff6f2 upstream

As newer glibcs will propagate the const attribute of the searched table
to its return.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/metricgroup.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/tools/perf/util/metricgroup.c b/tools/perf/util/metricgroup.c
index 4c98ac29ee13..664045fa0c8e 100644
--- a/tools/perf/util/metricgroup.c
+++ b/tools/perf/util/metricgroup.c
@@ -644,11 +644,10 @@ static const char *code_characters = ",-=@";
 
 static int encode_metric_id(struct strbuf *sb, const char *x)
 {
-	char *c;
 	int ret = 0;
 
 	for (; *x; x++) {
-		c = strchr(code_characters, *x);
+		const char *c = strchr(code_characters, *x);
 		if (c) {
 			ret = strbuf_addch(sb, '!');
 			if (ret)
-- 
2.54.0

[PATCH 6.1.y v2 11/18] perf session: Don't write to memory pointed to a const pointer

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit f1321cce848c558fde4c0c6bcd5e53f3cefd3af2 upstream

Since it is freshly allocated just attribute it to a non-const pointer
and then change it via that pointer.

That way we avoid const-correctness warnings in recent glibc versions.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/session.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index 1c4d124b1053..9d11cf2aab0c 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -2641,7 +2641,7 @@ bool perf_session__has_traces(struct perf_session *session, const char *msg)
 
 int map__set_kallsyms_ref_reloc_sym(struct map *map, const char *symbol_name, u64 addr)
 {
-	char *bracket;
+	char *bracket, *name;
 	struct ref_reloc_sym *ref;
 	struct kmap *kmap;
 
@@ -2649,13 +2649,13 @@ int map__set_kallsyms_ref_reloc_sym(struct map *map, const char *symbol_name, u6
 	if (ref == NULL)
 		return -ENOMEM;
 
-	ref->name = strdup(symbol_name);
+	ref->name = name = strdup(symbol_name);
 	if (ref->name == NULL) {
 		free(ref);
 		return -ENOMEM;
 	}
 
-	bracket = strchr(ref->name, ']');
+	bracket = strchr(name, ']');
 	if (bracket)
 		*bracket = '\0';
 
-- 
2.54.0

[PATCH 6.1.y v2 12/18] perf trace-event: Constify variables storing the result of strchr() on const tables

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit 97b81df7225830c4db3c17ed1235d2f3eb613d3d uptream

As newer glibcs will propagate the const attribute of the searched table
to its return.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/trace-event-info.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/trace-event-info.c b/tools/perf/util/trace-event-info.c
index 892c323b4ac9..5a6dec2dc0d8 100644
--- a/tools/perf/util/trace-event-info.c
+++ b/tools/perf/util/trace-event-info.c
@@ -477,7 +477,7 @@ static struct tracepoint_path *tracepoint_id_to_path(u64 config)
 static struct tracepoint_path *tracepoint_name_to_path(const char *name)
 {
 	struct tracepoint_path *path = zalloc(sizeof(*path));
-	char *str = strchr(name, ':');
+	const char *str = strchr(name, ':');
 
 	if (path == NULL || str == NULL) {
 		free(path);
-- 
2.54.0

[PATCH 6.1.y v2 13/18] perf units: Constify variables storing the result of strchr() on const tables

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit 0e14cb3b24f8f301cf6490a4493afc98321ed5bb upstream

As newer glibcs will propagate the const attribute of the searched table
to its return.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/units.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/units.c b/tools/perf/util/units.c
index 4c6a86e1cb54..0bbacf5a29aa 100644
--- a/tools/perf/util/units.c
+++ b/tools/perf/util/units.c
@@ -12,7 +12,7 @@ unsigned long parse_tag_value(const char *str, struct parse_tag *tags)
 	struct parse_tag *i = tags;
 
 	while (i->tag) {
-		char *s = strchr(str, i->tag);
+		const char *s = strchr(str, i->tag);
 
 		if (s) {
 			unsigned long int value;
-- 
2.54.0

[PATCH 6.1.y v2 14/18] perf bpf: Fix -Wdiscarded-qualifiers under C23

[Shreenidhi Shedi]

glibc >= 2.42 defaults to -std=gnu23, which promotes
    -Wdiscarded-qualifiers to an error.

util/bpf-loader.c: In function 'config_bpf_program':
util/bpf-loader.c:588:27: warning: initialization discards 'const' qualifier from pointer target type [-Wdiscarded-qualifiers]                                                                                                            588 |                 char *s = strchr(main_str, ':');

Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/bpf-loader.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/bpf-loader.c b/tools/perf/util/bpf-loader.c
index f4adeccdbbcb..7f4e774a3435 100644
--- a/tools/perf/util/bpf-loader.c
+++ b/tools/perf/util/bpf-loader.c
@@ -585,7 +585,7 @@ config_bpf_program(struct bpf_program *prog)
 		goto errout;
 
 	if (is_tp) {
-		char *s = strchr(main_str, ':');
+		const char *s = strchr(main_str, ':');
 
 		priv->is_tp = true;
 		priv->sys_name = strndup(main_str, s - main_str);
-- 
2.54.0

[PATCH 6.1.y v2 15/18] perf time-utils: Constify variables storing the result of strchr() on const tables

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit 21c0bc9144834e39762dd6fddbb255ebb80cf079 upstream

As newer glibcs will propagate the const attribute of the searched table
to its return.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/time-utils.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/time-utils.c b/tools/perf/util/time-utils.c
index 1b91ccd4d523..d43c4577d7eb 100644
--- a/tools/perf/util/time-utils.c
+++ b/tools/perf/util/time-utils.c
@@ -325,7 +325,7 @@ static int percent_comma_split(struct perf_time_interval *ptime_buf, int num,
 }
 
 static int one_percent_convert(struct perf_time_interval *ptime_buf,
-			       const char *ostr, u64 start, u64 end, char *c)
+			       const char *ostr, u64 start, u64 end, const char *c)
 {
 	char *str;
 	int len = strlen(ostr), ret;
@@ -358,7 +358,7 @@ static int one_percent_convert(struct perf_time_interval *ptime_buf,
 int perf_time__percent_parse_str(struct perf_time_interval *ptime_buf, int num,
 				 const char *ostr, u64 start, u64 end)
 {
-	char *c;
+	const char *c;
 
 	/*
 	 * ostr example:
-- 
2.54.0

[PATCH 6.1.y v2 16/18] perf demangle-java: Constify variables storing the result of strchr() on const tables

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit 79bba3a1834e7ba6c437674582cc9f3ae6fb638c upstream

As newer glibcs will propagate the const attribute of the searched table
to its return.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/demangle-java.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/demangle-java.c b/tools/perf/util/demangle-java.c
index ddf33d58bcd3..c3cb327ed562 100644
--- a/tools/perf/util/demangle-java.c
+++ b/tools/perf/util/demangle-java.c
@@ -158,7 +158,7 @@ char *
 java_demangle_sym(const char *str, int flags)
 {
 	char *buf, *ptr;
-	char *p;
+	const char *p;
 	size_t len, l1 = 0;
 
 	if (!str)
-- 
2.54.0

[PATCH 6.1.y v2 17/18] perf parse-events:: Fix -Wdiscarded-qualifiers under C23

[Shreenidhi Shedi]

glibc >= 2.42 defaults to -std=gnu23, which promotes
    -Wdiscarded-qualifiers to an error.

Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/parse-events.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c
index 5973f46c2375..e3b7950331bc 100644
--- a/tools/perf/util/parse-events.c
+++ b/tools/perf/util/parse-events.c
@@ -1647,7 +1647,7 @@ int parse_events__modifier_group(struct list_head *list,
  */
 static bool is_same_uncore_block(const char *pmu_name_a, const char *pmu_name_b)
 {
-	char *end_a, *end_b;
+	const char *end_a, *end_b;
 
 	end_a = strrchr(pmu_name_a, '_');
 	end_b = strrchr(pmu_name_b, '_');
-- 
2.54.0

[PATCH 6.1.y v2 18/18] perf bpf-event: Constify variables storing the result of strchr() on const tables

[Shreenidhi Shedi]

From: Arnaldo Carvalho de Melo <acme@redhat.com>

commit 8bf093acb3f1f07d846c86e32308f9f9954ed579 upstream

As newer glibcs will propagate the const attribute of the searched table
to its return.

Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: Shreenidhi Shedi <yesshedi@gmail.com>
---
 tools/perf/util/bpf-event.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/bpf-event.c b/tools/perf/util/bpf-event.c
index 91c7bfa82a50..fbaab3d3a476 100644
--- a/tools/perf/util/bpf-event.c
+++ b/tools/perf/util/bpf-event.c
@@ -411,7 +411,8 @@ kallsyms_process_symbol(void *data, const char *_name,
 			char type __maybe_unused, u64 start)
 {
 	char disp[KSYM_NAME_LEN];
-	char *module, *name;
+	const char *module;
+	char *name;
 	unsigned long id;
 	int err = 0;
 
-- 
2.54.0

Re: [PATCH 6.1.y v2 00/18] Backport fixes for -Wdiscarded-qualifiers and -Wnonnull with newer glibc

[Greg KH]

On Mon, May 11, 2026 at 12:40:33PM +0530, Shreenidhi Shedi wrote:
> Hi all,
> 
> This patch series backports a number of patches from master to 6.1.y
> to fix `-Wdiscarded-qualifiers` and `-Wnonnull` build issues with
> newer glibc versions.
> 
> I will port these changes to other stable trees once this gets reviewed.

You need to do this first for newer kernel trees, and only if they are
accepted there, should you do this for older ones as you do not want to
have regressions moving to newer kernels, right?

But first, why do this at all?  You should always be using the latest
kernel version of perf on older kernels, especially if you are updating
glibc.

And if you update glibc, WHY ARE YOU NOT UPDATING YOUR KERNEL?

Why would you be using an old kernel tree like this?  That's very odd,
please do not do that.

thanks,

greg k-h

Re: [PATCH 6.1.y v2 00/18] Backport fixes for -Wdiscarded-qualifiers and -Wnonnull with newer glibc

[Shreenidhi Shedi]

On 11/05/26 13:46, Greg KH wrote:
> On Mon, May 11, 2026 at 12:40:33PM +0530, Shreenidhi Shedi wrote:
>> Hi all,
>>
>> This patch series backports a number of patches from master to 6.1.y
>> to fix `-Wdiscarded-qualifiers` and `-Wnonnull` build issues with
>> newer glibc versions.
>>
>> I will port these changes to other stable trees once this gets reviewed.
> 
> You need to do this first for newer kernel trees, and only if they are
> accepted there, should you do this for older ones as you do not want to
> have regressions moving to newer kernels, right?
> 
> But first, why do this at all?  You should always be using the latest
> kernel version of perf on older kernels, especially if you are updating
> glibc.
> 
> And if you update glibc, WHY ARE YOU NOT UPDATING YOUR KERNEL?
> 
> Why would you be using an old kernel tree like this?  That's very odd,
> please do not do that.
> 
> thanks,
> 
> greg k-h

Hi Greg,

Thanks for the response. We have our own distro with 6.1.y kernel and we 
are trying to upgrade glibc to 2.43. As 6.1.y is well within support 
period I thought it would be good to keep it working with glibc-2.43 as 
these are harmless fixes and would help many (if someone is building 
6.1.y tree in Fedora rawhide for example).

Updating kernel to latest LTS is not feasible for us at the moment.

I will send a patch series to newer LTS releases soon. Thanks for the 
advice.

-- 
Shedi

Re: [PATCH 6.1.y v2 00/18] Backport fixes for -Wdiscarded-qualifiers and -Wnonnull with newer glibc

[Greg KH]

On Wed, May 13, 2026 at 12:07:30PM +0530, Shreenidhi Shedi wrote:
> On 11/05/26 13:46, Greg KH wrote:
> > On Mon, May 11, 2026 at 12:40:33PM +0530, Shreenidhi Shedi wrote:
> > > Hi all,
> > > 
> > > This patch series backports a number of patches from master to 6.1.y
> > > to fix `-Wdiscarded-qualifiers` and `-Wnonnull` build issues with
> > > newer glibc versions.
> > > 
> > > I will port these changes to other stable trees once this gets reviewed.
> > 
> > You need to do this first for newer kernel trees, and only if they are
> > accepted there, should you do this for older ones as you do not want to
> > have regressions moving to newer kernels, right?
> > 
> > But first, why do this at all?  You should always be using the latest
> > kernel version of perf on older kernels, especially if you are updating
> > glibc.
> > 
> > And if you update glibc, WHY ARE YOU NOT UPDATING YOUR KERNEL?
> > 
> > Why would you be using an old kernel tree like this?  That's very odd,
> > please do not do that.
> > 
> > thanks,
> > 
> > greg k-h
> 
> Hi Greg,
> 
> Thanks for the response. We have our own distro with 6.1.y kernel and we are
> trying to upgrade glibc to 2.43.

Great, then also update the kernel!  Don't only do one thing.

> As 6.1.y is well within support period I
> thought it would be good to keep it working with glibc-2.43 as these are
> harmless fixes and would help many (if someone is building 6.1.y tree in
> Fedora rawhide for example).

They shouldn't be doing that :)

> Updating kernel to latest LTS is not feasible for us at the moment.

Why not?  Nothing should break if you do that, just like nothing should
break when you update glibc.  What prevents you from doing this, you
will then get a few thousand CVEs fixed automatically for you that are
not currently fixed in the latest 6.1.y tree.

> I will send a patch series to newer LTS releases soon. Thanks for the
> advice.

No, this should not be needed, just build the latest perf if you really
need/want it.

thanks,

greg k-h