[PATCH 4/5] perf dso: Guard against cache underflow on short reads in dso_cache__memcpy()

[PATCH 4/5] perf dso: Guard against cache underflow on short reads in dso_cache__memcpy()

[Arnaldo Carvalho de Melo]

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

dso_cache__memcpy() computes cache_offset = offset - cache->offset,
then cache_size = min(cache->size - cache_offset, size).  The RB tree
lookup in __dso_cache__find() matches using the full
DSO__DATA_CACHE_SIZE window, but cache->size reflects the actual pread
return value from dso_cache__populate().

A short pread (e.g. near end-of-file) makes cache->size smaller than
DSO__DATA_CACHE_SIZE.  If a subsequent access targets an offset past
cache->offset + cache->size but within the DSO__DATA_CACHE_SIZE
window, the cache entry is found but cache_offset exceeds cache->size.
Since both are u64, the subtraction cache->size - cache_offset wraps
to a large value, min() selects the caller's size, and memcpy reads
out of bounds.

Return 0 (cache miss) when cache_offset falls outside the valid cached
range, so the caller re-reads from the backing file.

Fixes: 366df72657e0 ("perf dso: Refactor dso_cache__read()")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/dso.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 207f8744aac97e8c..a0de56c93592a5dd 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -1002,7 +1002,17 @@ static ssize_t dso_cache__memcpy(struct dso_cache *cache, u64 offset, u8 *data,
 				 u64 size, bool out)
 {
 	u64 cache_offset = offset - cache->offset;
-	u64 cache_size   = min(cache->size - cache_offset, size);
+	u64 cache_size;
+
+	/*
+	 * The RB tree matches using DSO__DATA_CACHE_SIZE, but a short
+	 * pread may leave cache->size smaller.  Treat an offset past
+	 * the valid data as a cache miss so the caller re-reads.
+	 */
+	if (cache_offset >= cache->size)
+		return 0;
+
+	cache_size = min(cache->size - cache_offset, size);
 
 	if (out)
 		memcpy(data, cache->data + cache_offset, cache_size);
-- 
2.55.0

Re: [PATCH 4/5] perf dso: Guard against cache underflow on short reads in dso_cache__memcpy()

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:

New issues:
- [High] Returning 0 from `dso_cache__memcpy()` causes a permanent EOF rather than triggering a cache miss and re-read as intended, breaking reads from dynamically growing files.

Pre-existing issues:
- [High] The RB-tree traversal in `__dso_cache__find()` is performed locklessly, creating a data race with concurrent cache insertions.
--

--- Patch [4]: [PATCH 4/5] perf dso: Guard against cache underflow on short reads in dso_cache__memcpy() ---
commit 06837abfd5cca9da5670ee54ec02e3164860900c
Author: Arnaldo Carvalho de Melo <acme@redhat.com>

perf dso: Guard against cache underflow on short reads in dso_cache__memcpy()

dso_cache__memcpy() computes cache_size based on cache->size, but a short
pread can leave cache->size smaller than the RB tree lookup window. This
causes a subtraction wrap-around and an out-of-bounds memcpy when accessed
past the cached size. This patch attempts to fix it by returning 0 (cache
miss) when the offset falls outside the valid cached range so the caller
re-reads from the backing file.

> diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
> index 207f8744aac97..a0de56c93592a 100644
> --- a/tools/perf/util/dso.c
> +++ b/tools/perf/util/dso.c
> @@ -1002,7 +1002,17 @@ static ssize_t dso_cache__memcpy(struct dso_cache *cache, u64 offset, u8 *data,
>  				 u64 size, bool out)
>  {
>  	u64 cache_offset = offset - cache->offset;
> -	u64 cache_size   = min(cache->size - cache_offset, size);
> +	u64 cache_size;
> +
> +	/*
> +	 * The RB tree matches using DSO__DATA_CACHE_SIZE, but a short
> +	 * pread may leave cache->size smaller.  Treat an offset past
> +	 * the valid data as a cache miss so the caller re-reads.
> +	 */
> +	if (cache_offset >= cache->size)
> +		return 0;

[Severity: High]
Does returning 0 here actually trigger a re-read?

Looking at cached_io() which calls this function, it seems to treat a 0 return
value strictly as an EOF condition:

cached_io() {
    ...
    ret = dso_cache__memcpy(cache, offset, data, size, out);
    if (!ret)
        break;
    ...
}

If the file has grown and we hit this condition, won't this cause cached_io()
to terminate the loop with a silent EOF instead of reading the new data from
the backing file?

> +
> +	cache_size = min(cache->size - cache_offset, size);
>  
>  	if (out)
>  		memcpy(data, cache->data + cache_offset, cache_size);

[Severity: High]
This is a pre-existing issue, but I noticed a potential data race when looking
at the RB-tree lookups in this file.

In __dso_cache__find(), the RB-tree traversal is performed locklessly:

    while (*p != NULL) {
        ...
        cache = rb_entry(parent, struct dso_cache, rb_node);
        ...
        if (offset < cache->offset)
            p = &(*p)->rb_left;
        else if (offset >= end)
            p = &(*p)->rb_right;
    }

However, dso_cache__insert() modifies the RB-tree while correctly holding
the dso__lock(dso) mutex:

    mutex_lock(dso__lock(dso));
    ...
    rb_link_node(&new->rb_node, parent, p);
    rb_insert_color(&new->rb_node, root);

If a lockless reader in __dso_cache__find() traverses the tree concurrently
with a cache miss that triggers dso_cache__insert(), could the reader observe
inconsistent tree pointers during the rb_insert_color() rotations?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260802142022.154219-1-acme@kernel.org?part=4

[PATCHES v2 0/5] perf DSO hardening series

[Arnaldo Carvalho de Melo]

Hi,

	Please consider merging,

Five hardening fixes for the perf DSO data handling code path, all found
by the sashiko-bot AI reviewer:

  - __open_dso() can return a file descriptor of 0 (stdin) when
    dso__get_filename() fails without setting errno, so fd = -errno = 0
    looked like a successfully opened DSO;

  - dso__decompress_kmodule_path() calls close(fd) with fd = -1 when
    decompression fails or the DSO is not compressed, clobbering errno
    with EBADF on the error path;

  - file_read() reads the stale global errno after try_to_open_dso()
    fails, when by then errno has been through mutex_lock(),
    nsinfo__mountns_enter() and several open() attempts;

  - dso_cache__memcpy() underflows when cache->size is smaller than the
    RB tree offset (short pread), computing a huge memcpy length;

  - dso__read_symbol() asserts on an input-dependent condition instead
    of doing a runtime check, keeping the BPF metadata safe from crafted
    input.

Best regards,

- Arnaldo

What changed from v1 (20260802142022.154219-1-acme@kernel.org):

  PATCH 1/5:
  - __open_dso() now sets errno = ENOENT directly when
    dso__get_filename() fails with errno == 0, instead of only computing
    fd = -ENOENT.  Callers that check errno after a negative fd (e.g.
    file_read() returning -errno) no longer get 0/EOF for an open
    failure [sashiko-bot review of PATCH 1/5].

  - Rebased onto the current perf-tools-next head (bf10e6ee2ac3034c).

This series was developed with assistance from Claude (claude-opus-4.6)
and Opencode (deepseek-v4-flash-free) for code analysis, patch
generation, and commit message composition.  All changes were validated
by the maintainer.

Arnaldo Carvalho de Melo (5):
  perf dso: Guard against errno==0 when dso__get_filename() returns NULL
  perf dso: Guard close() against invalid fd in
    dso__decompress_kmodule_path()
  perf dso: Use stored fd error instead of stale errno in file_read()
  perf dso: Guard against cache underflow on short reads in
    dso_cache__memcpy()
  perf dso: Replace assert with runtime check in dso__read_symbol()

 tools/perf/util/dso.c | 33 +++++++++++++++++++++++++++------
 1 file changed, 27 insertions(+), 6 deletions(-)

base-commit: bf10e6ee2ac3034c9068e03eed418fd16961984e
v1-head: f6cb9e46c7b8949b

[PATCH 1/5] perf dso: Guard against errno==0 when dso__get_filename() returns NULL

[Arnaldo Carvalho de Melo]

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

__open_dso() computes fd = -errno when dso__get_filename() returns NULL.
Some failure paths in dso__get_filename() (e.g. binary type mismatch)
return NULL without making a syscall, leaving errno at 0 from a prior
successful call.  fd = -0 = 0, which is stdin — subsequent code treats
it as a valid file descriptor.

Fall back to ENOENT when errno is 0, ensuring fd is always negative on
failure.

Fixes: eba5102d2f0b ("perf tools: Add global list of opened dso objects")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Cc: Jiri Olsa <jolsa@kernel.org>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/dso.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 2309196d8df3111c..fcd4b462992be6f2 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -640,10 +640,13 @@ static int __open_dso(struct dso *dso, struct machine *machine)
 	mutex_lock(dso__lock(dso));
 
 	name = dso__get_filename(dso, machine ? machine->root_dir : "", &decomp);
-	if (name)
+	if (name) {
 		fd = do_open(name);
-	else
+	} else {
+		if (errno == 0)
+			errno = ENOENT;
 		fd = -errno;
+	}
 
 	if (decomp)
 		unlink(name);
-- 
2.55.0

[PATCH 2/5] perf dso: Guard close() against invalid fd in dso__decompress_kmodule_path()

[Arnaldo Carvalho de Melo]

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

dso__decompress_kmodule_path() unconditionally calls close(fd) on the
return value of decompress_kmodule().  When decompression fails or the
DSO is not compressed, decompress_kmodule() returns -1.  close(-1)
fails with EBADF and clobbers errno, which callers up the chain
(dso__get_filename → __open_dso) depend on for error propagation.

Guard the close() call with fd >= 0 so only valid file descriptors are
closed.

Fixes: 42b3fa670825 ("perf tools: Introduce dso__decompress_kmodule_{fd,path}")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/dso.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index fcd4b462992be6f2..124193453675ca18 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -395,7 +395,9 @@ int dso__decompress_kmodule_path(struct dso *dso, const char *name,
 {
 	int fd = decompress_kmodule(dso, name, pathname, len);
 
-	close(fd);
+	/* decompress_kmodule() returns -1 on failure, don't close(-1) */
+	if (fd >= 0)
+		close(fd);
 	return fd >= 0 ? 0 : -1;
 }
 
-- 
2.55.0

[PATCH 3/5] perf dso: Use stored fd error instead of stale errno in file_read()

[Arnaldo Carvalho de Melo]

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

file_read() uses ret = -errno when dso__data(dso)->fd is negative after
try_to_open_dso() fails.  By this point errno has been through
mutex_lock(), nsinfo__mountns_enter(), and multiple open() attempts
inside try_to_open_dso() — it no longer reflects the actual open
failure.  If errno happens to be 0, ret = 0 looks like EOF rather than
an error.

The fd field already carries the negated errno from __open_dso()
(e.g. -EINVAL, -ENOENT), so use it directly instead of reading the
stale global errno.

Fixes: 33bdedcea2d7 ("perf tools: Protect dso cache fd with a mutex")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/dso.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 124193453675ca18..845a384a4d56779b 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -1029,7 +1029,8 @@ static ssize_t file_read(struct dso *dso, struct machine *machine,
 
 	if (dso__data(dso)->fd < 0) {
 		dso__data(dso)->status = DSO_DATA_STATUS_ERROR;
-		ret = -errno;
+		/* fd already carries the negated errno from __open_dso() */
+		ret = dso__data(dso)->fd;
 		goto out;
 	}
 
-- 
2.55.0

[PATCH 4/5] perf dso: Guard against cache underflow on short reads in dso_cache__memcpy()

[Arnaldo Carvalho de Melo]

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

dso_cache__memcpy() computes cache_offset = offset - cache->offset,
then cache_size = min(cache->size - cache_offset, size).  The RB tree
lookup in __dso_cache__find() matches using the full
DSO__DATA_CACHE_SIZE window, but cache->size reflects the actual pread
return value from dso_cache__populate().

A short pread (e.g. near end-of-file) makes cache->size smaller than
DSO__DATA_CACHE_SIZE.  If a subsequent access targets an offset past
cache->offset + cache->size but within the DSO__DATA_CACHE_SIZE
window, the cache entry is found but cache_offset exceeds cache->size.
Since both are u64, the subtraction cache->size - cache_offset wraps
to a large value, min() selects the caller's size, and memcpy reads
out of bounds.

Return 0 (cache miss) when cache_offset falls outside the valid cached
range, so the caller re-reads from the backing file.

Fixes: 366df72657e0 ("perf dso: Refactor dso_cache__read()")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/dso.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 845a384a4d56779b..2cf9f44a87903d7d 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -1005,7 +1005,17 @@ static ssize_t dso_cache__memcpy(struct dso_cache *cache, u64 offset, u8 *data,
 				 u64 size, bool out)
 {
 	u64 cache_offset = offset - cache->offset;
-	u64 cache_size   = min(cache->size - cache_offset, size);
+	u64 cache_size;
+
+	/*
+	 * The RB tree matches using DSO__DATA_CACHE_SIZE, but a short
+	 * pread may leave cache->size smaller.  Treat an offset past
+	 * the valid data as a cache miss so the caller re-reads.
+	 */
+	if (cache_offset >= cache->size)
+		return 0;
+
+	cache_size = min(cache->size - cache_offset, size);
 
 	if (out)
 		memcpy(data, cache->data + cache_offset, cache_size);
-- 
2.55.0

[PATCH 5/5] perf dso: Replace assert with runtime check in dso__read_symbol()

[Arnaldo Carvalho de Melo]

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

dso__read_symbol() asserts that len <= jited_prog_len, where len comes
from sym->end - sym->start (parsed from PERF_RECORD_KSYMBOL in
perf.data).  Both values originate from untrusted file input.

With NDEBUG (production builds), the assert is compiled out, allowing
an out-of-bounds heap read when the BPF program buffer is accessed.
Without NDEBUG, a crafted perf.data crashes perf with an assertion
failure.

Replace the assert with a runtime bounds check that returns NULL with
an appropriate error code, matching the existing error handling
pattern in this function.

Fixes: aa04707f507e ("perf dso: Support BPF programs in dso__read_symbol()")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Cc: Ian Rogers <irogers@google.com>
Cc: Song Liu <song@kernel.org>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/dso.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 2cf9f44a87903d7d..031de8f1da0329b2 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -2027,7 +2027,12 @@ const u8 *dso__read_symbol(struct dso *dso, const char *symfs_filename,
 			errno = SYMBOL_ANNOTATE_ERRNO__BPF_MISSING_BTF;
 			return NULL;
 		}
-		assert(len <= info_linear->info.jited_prog_len);
+		if (len > info_linear->info.jited_prog_len) {
+			pr_debug("BPF symbol length %zu exceeds jited_prog_len %u\n",
+				 len, info_linear->info.jited_prog_len);
+			errno = SYMBOL_ANNOTATE_ERRNO__BPF_MISSING_BTF;
+			return NULL;
+		}
 		*out_buf_len = len;
 		return (const u8 *)(uintptr_t)(info_linear->info.jited_prog_insns);
 #else
-- 
2.55.0

[PATCH 4/5] perf dso: Guard against cache underflow on short reads in dso_cache__memcpy()

[Arnaldo Carvalho de Melo]

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

dso_cache__memcpy() computes cache_offset = offset - cache->offset,
then cache_size = min(cache->size - cache_offset, size).  The RB tree
lookup in __dso_cache__find() matches using the full
DSO__DATA_CACHE_SIZE window, but cache->size reflects the actual pread
return value from dso_cache__populate().

A short pread (e.g. near end-of-file) makes cache->size smaller than
DSO__DATA_CACHE_SIZE.  If a subsequent access targets an offset past
cache->offset + cache->size but within the DSO__DATA_CACHE_SIZE
window, the cache entry is found but cache_offset exceeds cache->size.
Since both are u64, the subtraction cache->size - cache_offset wraps
to a large value, min() selects the caller's size, and memcpy reads
out of bounds.

Return 0 for an offset past the valid cached data.  For a regular
file a short pread only happens at end-of-file, so 0 is what a direct
pread() at that offset would return: cached_io() stops its read loop
as on EOF.  Re-reading from the backing file would not help — a
second pread at the same offset returns the same short count.

Fixes: 366df72657e0 ("perf dso: Refactor dso_cache__read()")
Reported-by: sashiko-bot <sashiko-bot@kernel.org>
Cc: Adrian Hunter <adrian.hunter@intel.com>
Assisted-by: Claude:claude-opus-4.6
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/util/dso.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/dso.c b/tools/perf/util/dso.c
index 32ae5c78cdb906e6..8e16b919e80721f0 100644
--- a/tools/perf/util/dso.c
+++ b/tools/perf/util/dso.c
@@ -1005,7 +1005,20 @@ static ssize_t dso_cache__memcpy(struct dso_cache *cache, u64 offset, u8 *data,
 				 u64 size, bool out)
 {
 	u64 cache_offset = offset - cache->offset;
-	u64 cache_size   = min(cache->size - cache_offset, size);
+	u64 cache_size;
+
+	/*
+	 * The RB tree matches using DSO__DATA_CACHE_SIZE, but a short
+	 * pread may leave cache->size smaller.  For a regular file a
+	 * short pread only happens at end-of-file, so an offset past
+	 * the valid data is EOF: return 0, matching what a direct
+	 * pread() at that offset would return, and cached_io() then
+	 * stops its read loop.
+	 */
+	if (cache_offset >= cache->size)
+		return 0;
+
+	cache_size = min(cache->size - cache_offset, size);
 
 	if (out)
 		memcpy(data, cache->data + cache_offset, cache_size);
-- 
2.55.0