Skip to contentSkip to conversation

bpf: Implement stack_map_get_build_id_offset_sleepable()

PATCH Applied Ihor Solodrai 5 participants 10 messages May 25 – Jun 3

Subsystems (2)
lore
Share & tools
lore

Ihor Solodrai

author reviewer
#0 lore raw

v7 cover bpf: Implement stack_map_get_build_id_offset_sleepable()

The series introduces stack_map_get_build_id_offset_sleepable(), fixing a gap with parsing build_id in sleepable context in stackmap.c

In particular, this fixes a deadlock in stack_map_get_build_id_offset() doing a blocking __kernel_read(), which happens since commit 777a8560fd29 ("lib/buildid: use __kernel_read() for sleepable context").

See previous revisions for more details.

---

v6->v7:
* Addressed feedback from Andrii (mostly patch #2):
* implement proper CONFIG_PER_VMA_LOCK=n support, following a VMA locking pattern similar to one used in PROCMAP_QUERY
* change the contract of stack_map_lock_vma(): if a non-NULL VMA is returned, then a read lock is held
* remove now unnecessary vma_locked flag
* and various other nits
* Add vma_is_anonymous() checks where appropriate (AIs)
v6: https://lore.kernel.org/bpf/[email protected]/

v5->v6:
* Misc refactoring (Andrii):
* add stack_map_build_id_set_valid() helper
* simplify control flow in stack_map_get_build_id_offset_sleepable()
v5: https://lore.kernel.org/bpf/[email protected]/

v4->v5:
* Add comments explaining mmap_read_trylock() (Shakeel)
* Rebase on bpf-next (Alexei)
v4: https://lore.kernel.org/bpf/[email protected]/

v3->v4:
* Change Fixes tag in patch #2 (AI)
* Nit in caching implementation (Mykyta)
v3: https://lore.kernel.org/bpf/[email protected]/

v2->v3:
* Split patch #2 in two: stack_map_get_build_id_offset_sleepable() implementation, and then introduce caching
* Drop taking mmap_lock if CONFIG_PER_VMA_LOCK=n, fall back to raw IPs instead
* Cache vm_{start,end} in addition to prev_file (Mykyta)
v2: https://lore.kernel.org/bpf/[email protected]/

v1->v2:
* Addressed feedback from Puranjay:
* split out a small refactoring patch
* use mmap_read_trylock()
* take into account CONFIG_PER_VMA_LOCK
* replace find_vma() with vma_lookup()
* cache prev_build_id to avoid re-parsing the same file
* Snapshot vm_pgoff and vm_start before unlocking (AI)
* To avoid repetitive unlocking statements, introduce struct stack_map_vma_lock to hold relevant lock state info and add an unlock helper
v1: https://lore.kernel.org/bpf/[email protected]/

---

Ihor Solodrai (3):
bpf: Factor out stack_map build ID helpers
bpf: Avoid faultable build ID reads under mm locks
bpf: Cache build IDs in sleepable stackmap path

 kernel/bpf/stackmap.c | 215 ++++++++++++++++++++++++++++++++++++++----
 1 file changed, 198 insertions(+), 17 deletions(-)
Signature · 1 line
2.54.0
author reviewer
#1 lore raw

PATCH v7 1/3 bpf: Factor out stack_map build ID helpers

Factor out helpers from stack_map_get_build_id_offset() in preparation for adding a sleepable build ID resolution path:
stack_map_build_id_set_ip(), stack_map_build_id_offset(), and stack_map_build_id_set_valid().

While here, refactor stack_map_get_build_id_offset():
* use continue-driven control flow in the main loop and remove build_id_valid label
* update prev_vma and prev_build_id on the fall-back-to-IP branch so the cache reflects the actual VMA seen on the previous IP [1]
* guard fetch_build_id() with vma_is_anonymous() [2] to skip parse attempts that would otherwise fail the ELF magic check

[1] https://lore.kernel.org/bpf/CAEf4Bzac9uWWqBvzH0iFzKvJcq3vxscZ3pKm0sUHmN-F-z9wVQ@mail.gmail.com/
[2] https://lore.kernel.org/bpf/226398c1ff3f2b686c0aeb010408d85fb15df13f9ff60a045bee31e79b9e41e9@mail.kernel.org/

---
 kernel/bpf/stackmap.c | 57 ++++++++++++++++++++++++++++++-------------
 1 file changed, 40 insertions(+), 17 deletions(-)
kernel/bpf/stackmap.c +40 -17
index da3d328f5c15..e23be7d44503 100644
@@ -152,6 +152,28 @@ static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b
 			 : build_id_parse_nofault(vma, build_id, NULL);
 }
 
+static inline void stack_map_build_id_set_ip(struct bpf_stack_build_id *id)
+{
+	id->status = BPF_STACK_BUILD_ID_IP;
+	memset(id->build_id, 0, BUILD_ID_SIZE_MAX);
+}
+
+static inline u64 stack_map_build_id_offset(unsigned long vm_pgoff,
+					    unsigned long vm_start, u64 ip)
+{
+	return (vm_pgoff << PAGE_SHIFT) + ip - vm_start;
+}
+
+static inline void stack_map_build_id_set_valid(struct bpf_stack_build_id *id,
+						u64 offset,
+						const unsigned char *build_id)
+{
+	id->status = BPF_STACK_BUILD_ID_VALID;
+	id->offset = offset;
+	if (id->build_id != build_id)
+		memcpy(id->build_id, build_id, BUILD_ID_SIZE_MAX);
+}
+
 /*
  * Expects all id_offs[i].ip values to be set to correct initial IPs.
  * They will be subsequently:
@@ -165,44 +187,45 @@ static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b
 static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs,
 					  u32 trace_nr, bool user, bool may_fault)
 {
-	int i;
 	struct mmap_unlock_irq_work *work = NULL;
 	bool irq_work_busy = bpf_mmap_unlock_get_irq_work(&work);
+	bool has_user_ctx = user && current && current->mm;
 	struct vm_area_struct *vma, *prev_vma = NULL;
-	const char *prev_build_id;
+	const unsigned char *prev_build_id = NULL;
+	int i;
 
 	/* If the irq_work is in use, fall back to report ips. Same
 	 * fallback is used for kernel stack (!user) on a stackmap with
 	 * build_id.
 	 */
-	if (!user || !current || !current->mm || irq_work_busy ||
-	    !mmap_read_trylock(current->mm)) {
+	if (!has_user_ctx || irq_work_busy || !mmap_read_trylock(current->mm)) {
 		/* cannot access current->mm, fall back to ips */
-		for (i = 0; i < trace_nr; i++) {
-			id_offs[i].status = BPF_STACK_BUILD_ID_IP;
-			memset(id_offs[i].build_id, 0, BUILD_ID_SIZE_MAX);
-		}
+		for (i = 0; i < trace_nr; i++)
+			stack_map_build_id_set_ip(&id_offs[i]);
 		return;
 	}
 
 	for (i = 0; i < trace_nr; i++) {
 		u64 ip = READ_ONCE(id_offs[i].ip);
+		u64 offset;
 
-		if (range_in_vma(prev_vma, ip, ip)) {
+		if (prev_build_id && range_in_vma(prev_vma, ip, ip)) {
 			vma = prev_vma;
-			memcpy(id_offs[i].build_id, prev_build_id, BUILD_ID_SIZE_MAX);
-			goto build_id_valid;
+			offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+			stack_map_build_id_set_valid(&id_offs[i], offset, prev_build_id);
+			continue;
 		}
 		vma = find_vma(current->mm, ip);
-		if (!vma || fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
+		if (!vma || vma_is_anonymous(vma) ||
+		    fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
 			/* per entry fall back to ips */
-			id_offs[i].status = BPF_STACK_BUILD_ID_IP;
-			memset(id_offs[i].build_id, 0, BUILD_ID_SIZE_MAX);
+			stack_map_build_id_set_ip(&id_offs[i]);
+			prev_vma = vma;
+			prev_build_id = NULL;
 			continue;
 		}
-build_id_valid:
-		id_offs[i].offset = (vma->vm_pgoff << PAGE_SHIFT) + ip - vma->vm_start;
-		id_offs[i].status = BPF_STACK_BUILD_ID_VALID;
+		offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+		stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
 		prev_vma = vma;
 		prev_build_id = id_offs[i].build_id;
 	}
Signature · 1 line
2.54.0
#2 lore raw

Re: [PATCH bpf-next v7 1/3] bpf: Factor out stack_map build ID helpers

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Medium] The refactored VMA caching logic bypasses the cache for fallback cases and evicts valid VMAs…

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The refactored VMA caching logic bypasses the cache for fallback cases and evicts valid VMAs, directly contradicting the patch's stated goal. --

commit e38559ed451d61a994765f3bfb260cc14edd3569
Author: Ihor Solodrai <[email protected]>

bpf: Factor out stack_map build ID helpers

This patch refactors stack_map_get_build_id_offset() to use helpers and continue-driven control flow. It also attempts to optimize VMA caching by updating the cached VMA on the fallback-to-IP branch to prevent redundant parsing attempts for anonymous or failing VMAs.

Quoted 4 lines
kernel/bpf/stackmap.c +0 -0
index da3d328f5c15a..e23be7d445038 100644

[ ... ]

Quoted 4 lines
kernel/bpf/stackmap.c +0 -0
@@ -165,44 +187,45 @@ static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b
 static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs,
 					  u32 trace_nr, bool user, bool may_fault)
 {

[ ... ]

Quoted 13 lines
kernel/bpf/stackmap.c +5 -3
 	for (i = 0; i < trace_nr; i++) {
 		u64 ip = READ_ONCE(id_offs[i].ip);
+		u64 offset;
 
-		if (range_in_vma(prev_vma, ip, ip)) {
+		if (prev_build_id && range_in_vma(prev_vma, ip, ip)) {
 			vma = prev_vma;
-			memcpy(id_offs[i].build_id, prev_build_id, BUILD_ID_SIZE_MAX);
-			goto build_id_valid;
+			offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+			stack_map_build_id_set_valid(&id_offs[i], offset, prev_build_id);
+			continue;
 		}

kernel/bpf/stackmap.c static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b · line 224

Quoted 10 lines
kernel/bpf/stackmap.c +5 -3
 		vma = find_vma(current->mm, ip);
-		if (!vma || fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
+		if (!vma || vma_is_anonymous(vma) ||
+		    fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
 			/* per entry fall back to ips */
-			id_offs[i].status = BPF_STACK_BUILD_ID_IP;
-			memset(id_offs[i].build_id, 0, BUILD_ID_SIZE_MAX);
+			stack_map_build_id_set_ip(&id_offs[i]);
+			prev_vma = vma;
+			prev_build_id = NULL;

[Severity: Medium]
Does setting prev_build_id to NULL here inadvertently defeat the VMA caching optimization mentioned in the commit message?

The commit message states the intent is to update prev_vma and prev_build_id on this fallback branch so the cache reflects the actual VMA seen. However, the cache check at the top of the loop now requires prev_build_id to be non-NULL:

    if (prev_build_id && range_in_vma(prev_vma, ip, ip)) {

Since prev_build_id is set to NULL on failure, won't consecutive IPs falling into the same anonymous or failing VMA bypass the cache completely and trigger redundant find_vma() lookups?

Also, does overwriting prev_vma with the failing VMA while disabling the cache hit (via NULL prev_build_id) evict the last successful VMA from the cache, potentially degrading performance if stack frames interleave?

Quoted 10 lines
kernel/bpf/stackmap.c +3 -3
+			continue;
 		}
-build_id_valid:
-		id_offs[i].offset = (vma->vm_pgoff << PAGE_SHIFT) + ip - vma->vm_start;
-		id_offs[i].status = BPF_STACK_BUILD_ID_VALID;
+		offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+		stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
 		prev_vma = vma;
 		prev_build_id = id_offs[i].build_id;
 	}
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1
#3 lore raw

Re: [PATCH bpf-next v7 1/3] bpf: Factor out stack_map build ID helpers

Quoted 65 lines · On Mon, May 25, 2026 at 4:11 PM <[email protected]> wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The refactored VMA caching logic bypasses the cache for fallback cases and evicts valid VMAs, directly contradicting the patch's stated goal. --

commit e38559ed451d61a994765f3bfb260cc14edd3569
Author: Ihor Solodrai <[email protected]>

bpf: Factor out stack_map build ID helpers

This patch refactors stack_map_get_build_id_offset() to use helpers and continue-driven control flow. It also attempts to optimize VMA caching by updating the cached VMA on the fallback-to-IP branch to prevent redundant parsing attempts for anonymous or failing VMAs.

Quoted 4 lines
kernel/bpf/stackmap.c +0 -0
index da3d328f5c15a..e23be7d445038 100644

[ ... ]

Quoted 4 lines
kernel/bpf/stackmap.c +0 -0
@@ -165,44 +187,45 @@ static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b
 static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs,
                                        u32 trace_nr, bool user, bool may_fault)
 {

[ ... ]

Quoted 13 lines
      for (i = 0; i < trace_nr; i++) {
              u64 ip = READ_ONCE(id_offs[i].ip);
kernel/bpf/stackmap.c +5 -3
+             u64 offset;
 
-             if (range_in_vma(prev_vma, ip, ip)) {
+             if (prev_build_id && range_in_vma(prev_vma, ip, ip)) {
                      vma = prev_vma;
-                     memcpy(id_offs[i].build_id, prev_build_id, BUILD_ID_SIZE_MAX);
-                     goto build_id_valid;
+                     offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+                     stack_map_build_id_set_valid(&id_offs[i], offset, prev_build_id);
+                     continue;
              }

kernel/bpf/stackmap.c static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b · line 224

Quoted 10 lines
kernel/bpf/stackmap.c +5 -3
              vma = find_vma(current->mm, ip);
-             if (!vma || fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
+             if (!vma || vma_is_anonymous(vma) ||
+                 fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
                      /* per entry fall back to ips */
-                     id_offs[i].status = BPF_STACK_BUILD_ID_IP;
-                     memset(id_offs[i].build_id, 0, BUILD_ID_SIZE_MAX);
+                     stack_map_build_id_set_ip(&id_offs[i]);
+                     prev_vma = vma;
+                     prev_build_id = NULL;

[Severity: Medium]
Does setting prev_build_id to NULL here inadvertently defeat the VMA caching optimization mentioned in the commit message?

The commit message states the intent is to update prev_vma and prev_build_id on this fallback branch so the cache reflects the actual VMA seen. However, the cache check at the top of the loop now requires prev_build_id to be non-NULL:

Quoted 5 lines
    if (prev_build_id && range_in_vma(prev_vma, ip, ip)) {

Since prev_build_id is set to NULL on failure, won't consecutive IPs falling into the same anonymous or failing VMA bypass the cache completely and trigger redundant find_vma() lookups?

yeah, it seems so to me... we have prev_build_id and prev_vma, we should set prev_build_id to NULL if build ID fetching failed, but we should preserve prev_vma always to avoid unnecessary find_vma() calls. And if it so happens that we have matching prev_vma, then depending of prev_build_id being NULL or not we either set absolute ip or build_id+offset, respectively

Let's do a follow up to improve this, nothing is really broken, I think, so applying.

Quoted 19 lines

Also, does overwriting prev_vma with the failing VMA while disabling the cache hit (via NULL prev_build_id) evict the last successful VMA from the cache, potentially degrading performance if stack frames interleave?

kernel/bpf/stackmap.c static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b · line 231

Quoted 10 lines
kernel/bpf/stackmap.c +3 -3
+                     continue;
              }
-build_id_valid:
-             id_offs[i].offset = (vma->vm_pgoff << PAGE_SHIFT) + ip - vma->vm_start;
-             id_offs[i].status = BPF_STACK_BUILD_ID_VALID;
+             offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+             stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
              prev_vma = vma;
              prev_build_id = id_offs[i].build_id;
      }

--
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1

author reviewer
#4 lore raw

Re: [PATCH bpf-next v7 1/3] bpf: Factor out stack_map build ID helpers

Quoted 79 lines · Andrii Nakryiko
Quoted 65 lines · On Mon, May 25, 2026 at 4:11 PM <[email protected]> wrote:

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The refactored VMA caching logic bypasses the cache for fallback cases and evicts valid VMAs, directly contradicting the patch's stated goal. --

commit e38559ed451d61a994765f3bfb260cc14edd3569
Author: Ihor Solodrai <[email protected]>

bpf: Factor out stack_map build ID helpers

This patch refactors stack_map_get_build_id_offset() to use helpers and continue-driven control flow. It also attempts to optimize VMA caching by updating the cached VMA on the fallback-to-IP branch to prevent redundant parsing attempts for anonymous or failing VMAs.

Quoted 4 lines
kernel/bpf/stackmap.c +0 -0
index da3d328f5c15a..e23be7d445038 100644

[ ... ]

Quoted 4 lines
kernel/bpf/stackmap.c +0 -0
@@ -165,44 +187,45 @@ static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b
  static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs,
                                         u32 trace_nr, bool user, bool may_fault)
  {

[ ... ]

Quoted 13 lines
       for (i = 0; i < trace_nr; i++) {
               u64 ip = READ_ONCE(id_offs[i].ip);
kernel/bpf/stackmap.c +5 -3
+             u64 offset;
 
-             if (range_in_vma(prev_vma, ip, ip)) {
+             if (prev_build_id && range_in_vma(prev_vma, ip, ip)) {
                       vma = prev_vma;
-                     memcpy(id_offs[i].build_id, prev_build_id, BUILD_ID_SIZE_MAX);
-                     goto build_id_valid;
+                     offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+                     stack_map_build_id_set_valid(&id_offs[i], offset, prev_build_id);
+                     continue;
               }

kernel/bpf/stackmap.c static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b · line 224

Quoted 10 lines
kernel/bpf/stackmap.c +5 -3
               vma = find_vma(current->mm, ip);
-             if (!vma || fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
+             if (!vma || vma_is_anonymous(vma) ||
+                 fetch_build_id(vma, id_offs[i].build_id, may_fault)) {
                       /* per entry fall back to ips */
-                     id_offs[i].status = BPF_STACK_BUILD_ID_IP;
-                     memset(id_offs[i].build_id, 0, BUILD_ID_SIZE_MAX);
+                     stack_map_build_id_set_ip(&id_offs[i]);
+                     prev_vma = vma;
+                     prev_build_id = NULL;

[Severity: Medium]
Does setting prev_build_id to NULL here inadvertently defeat the VMA caching optimization mentioned in the commit message?

The commit message states the intent is to update prev_vma and prev_build_id on this fallback branch so the cache reflects the actual VMA seen. However, the cache check at the top of the loop now requires prev_build_id to be non-NULL:

Quoted 5 lines
     if (prev_build_id && range_in_vma(prev_vma, ip, ip)) {

Since prev_build_id is set to NULL on failure, won't consecutive IPs falling into the same anonymous or failing VMA bypass the cache completely and trigger redundant find_vma() lookups?

yeah, it seems so to me... we have prev_build_id and prev_vma, we should set prev_build_id to NULL if build ID fetching failed, but we should preserve prev_vma always to avoid unnecessary find_vma() calls. And if it so happens that we have matching prev_vma, then depending of prev_build_id being NULL or not we either set absolute ip or build_id+offset, respectively

Quoted 2 lines

Let's do a follow up to improve this, nothing is really broken, I think, so applying.

Hi Andrii, thanks for applying.

I'll send a follow up as soon as I can.

Quoted 21 lines
Quoted 21 lines

Also, does overwriting prev_vma with the failing VMA while disabling the cache hit (via NULL prev_build_id) evict the last successful VMA from the cache, potentially degrading performance if stack frames interleave?

kernel/bpf/stackmap.c static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b · line 231

Quoted 10 lines
kernel/bpf/stackmap.c +3 -3
+                     continue;
               }
-build_id_valid:
-             id_offs[i].offset = (vma->vm_pgoff << PAGE_SHIFT) + ip - vma->vm_start;
-             id_offs[i].status = BPF_STACK_BUILD_ID_VALID;
+             offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+             stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
               prev_vma = vma;
               prev_build_id = id_offs[i].build_id;
       }

--
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=1

#5 lore raw

Re: [PATCH bpf-next v7 1/3] bpf: Factor out stack_map build ID helpers

In reply to Ihor Solodrai
In bpf_get_stack(), the destination buffer can be a BPF map value that is concurrently readable locklessly by user-space. Since the buffer initially contains the raw instruction pointer (ip), does…
Quoted 19 lines
kernel/bpf/stackmap.c +11 -0
index da3d328f5c15..e23be7d44503 100644
@@ -152,6 +152,28 @@ static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b
 			 : build_id_parse_nofault(vma, build_id, NULL);
 }
 
+static inline void stack_map_build_id_set_ip(struct bpf_stack_build_id *id)
+{
+	id->status = BPF_STACK_BUILD_ID_IP;
+	memset(id->build_id, 0, BUILD_ID_SIZE_MAX);
+}
+
+static inline u64 stack_map_build_id_offset(unsigned long vm_pgoff,
+					    unsigned long vm_start, u64 ip)
+{
+	return (vm_pgoff << PAGE_SHIFT) + ip - vm_start;
+}

kernel/bpf/stackmap.c static int fetch_build_id(struct vm_area_struct *vma, unsigned char *build_id, b · line 175

Quoted 10 lines
kernel/bpf/stackmap.c +10 -0
+
+static inline void stack_map_build_id_set_valid(struct bpf_stack_build_id *id,
+						u64 offset,
+						const unsigned char *build_id)
+{
+	id->status = BPF_STACK_BUILD_ID_VALID;
+	id->offset = offset;
+	if (id->build_id != build_id)
+		memcpy(id->build_id, build_id, BUILD_ID_SIZE_MAX);
+}

In bpf_get_stack(), the destination buffer can be a BPF map value that is concurrently readable locklessly by user-space. Since the buffer initially contains the raw instruction pointer (ip), does setting status to VALID before writing the offset create a data race?

A concurrent lockless reader polling the status field might observe the VALID status while the union still holds the raw ip instead of the correct file offset, and before the build_id string is fully copied.

Prior to this patch, the offset was written before the status. Could this reordering of assignments cause concurrent readers to observe stale data?

This was raised by sashiko-bot in v6 review and acknowledged by the author as needing to be fixed in the next version, but was not addressed:

https://lore.kernel.org/bpf/[email protected]/

[ ... ]

---
AI reviewed your patch. Please fix the bug or email reply why it's not a bug.
See: https://github.com/kernel-patches/vmtest/blob/master/ci/claude/README.md

CI run summary: https://github.com/kernel-patches/bpf/actions/runs/26422948694

author reviewer
#6 lore raw

PATCH v7 2/3 bpf: Avoid faultable build ID reads under mm locks

In reply to Ihor Solodrai

Sleepable build ID parsing can block in __kernel_read() [1], so the stackmap sleepable path must not call it while holding mmap_lock or a per-VMA read lock.

The issue and the fix are conceptually similar to a recent procfs patch [2]. A similar VMA locking pattern has already been used in PROCMAP_QUERY [3].

Resolve each covered VMA with a stable read-side reference, preferring lock_vma_under_rcu() and falling back to mmap_read_trylock() only long enough to acquire the VMA read lock. Take a reference to the backing file, drop the VMA lock, and then parse the build ID through (sleepable) build_id_parse_file().

We have to use mmap_read_trylock() (and give up on failure) in this context because taking mmap_read_lock() is generally unsafe on code paths reachable from BPF programs [4], and may lead to deadlocks.

[1] https://lore.kernel.org/all/[email protected]/
[2] https://lore.kernel.org/all/[email protected]/
[3] https://lore.kernel.org/all/[email protected]/
[4] https://lore.kernel.org/bpf/[email protected]/

Fixes d4dd9775ec24 ("bpf: wire up sleepable bpf_get_stack() and bpf_get_task_stack() helpers")
---
 kernel/bpf/stackmap.c | 109 ++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 109 insertions(+)
kernel/bpf/stackmap.c +109 -0
index e23be7d44503..c53cfd9a67cf 100644
@@ -9,6 +9,7 @@
 #include <linux/perf_event.h>
 #include <linux/btf_ids.h>
 #include <linux/buildid.h>
+#include <linux/mmap_lock.h>
 #include "percpu_freelist.h"
 #include "mmap_unlock_work.h"
 
@@ -174,6 +175,109 @@ static inline void stack_map_build_id_set_valid(struct bpf_stack_build_id *id,
 		memcpy(id->build_id, build_id, BUILD_ID_SIZE_MAX);
 }
 
+struct stack_map_vma_lock {
+	struct vm_area_struct *vma;
+	struct mm_struct *mm;
+};
+
+/*
+ * Acquire a stable read-side reference on the VMA covering @ip.
+ *
+ * With CONFIG_PER_VMA_LOCK=y this returns a VMA with its per-VMA read
+ * lock held and mmap_lock dropped, so the caller may sleep.
+ *
+ * With CONFIG_PER_VMA_LOCK=n it returns a VMA with mmap_lock still
+ * held; the caller must snapshot any fields it needs and pin vm_file
+ * with get_file() before stack_map_unlock_vma() drops mmap_lock, as
+ * the VMA may be split, merged, or freed after that.
+ *
+ * Returns NULL on failure, in which case no lock is held.
+ */
+static struct vm_area_struct *
+stack_map_lock_vma(struct stack_map_vma_lock *lock, unsigned long ip)
+{
+	struct mm_struct *mm = lock->mm;
+	struct vm_area_struct *vma;
+
+	/* noop under !CONFIG_PER_VMA_LOCK */
+	vma = lock_vma_under_rcu(mm, ip);
+	if (vma) {
+		lock->vma = vma;
+		return vma;
+	}
+
+	/*
+	 * Taking mmap_read_lock() is unsafe here, because the caller BPF
+	 * program might already hold it, causing a deadlock.
+	 */
+	if (!mmap_read_trylock(mm))
+		return NULL;
+
+	vma = vma_lookup(mm, ip);
+	if (!vma) {
+		mmap_read_unlock(mm);
+		return NULL;
+	}
+
+#ifdef CONFIG_PER_VMA_LOCK
+	if (!vma_start_read_locked(vma)) {
+		mmap_read_unlock(mm);
+		return NULL;
+	}
+	mmap_read_unlock(mm);
+#endif
+
+	lock->vma = vma;
+	return vma;
+}
+
+static void stack_map_unlock_vma(struct stack_map_vma_lock *lock)
+{
+#ifdef CONFIG_PER_VMA_LOCK
+	vma_end_read(lock->vma);
+#else
+	mmap_read_unlock(lock->mm);
+#endif
+	lock->vma = NULL;
+}
+
+static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *id_offs,
+						    u32 trace_nr)
+{
+	struct mm_struct *mm = current->mm;
+	struct stack_map_vma_lock lock = { .mm = mm };
+	struct vm_area_struct *vma;
+	struct file *file;
+	u64 offset;
+	u64 ip;
+
+	for (u32 i = 0; i < trace_nr; i++) {
+		ip = READ_ONCE(id_offs[i].ip);
+
+		vma = stack_map_lock_vma(&lock, ip);
+		if (!vma) {
+			stack_map_build_id_set_ip(&id_offs[i]);
+			continue;
+		}
+		if (vma_is_anonymous(vma) || !vma->vm_file) {
+			stack_map_build_id_set_ip(&id_offs[i]);
+			stack_map_unlock_vma(&lock);
+			continue;
+		}
+
+		file = get_file(vma->vm_file);
+		offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+		stack_map_unlock_vma(&lock);
+
+		/* build_id_parse_file() may block on filesystem reads */
+		if (build_id_parse_file(file, id_offs[i].build_id, NULL))
+			stack_map_build_id_set_ip(&id_offs[i]);
+		else
+			stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
+		fput(file);
+	}
+}
+
 /*
  * Expects all id_offs[i].ip values to be set to correct initial IPs.
  * They will be subsequently:
@@ -194,6 +298,11 @@ static void stack_map_get_build_id_offset(struct bpf_stack_build_id *id_offs,
 	const unsigned char *prev_build_id = NULL;
 	int i;
 
+	if (may_fault && has_user_ctx) {
+		stack_map_get_build_id_offset_sleepable(id_offs, trace_nr);
+		return;
+	}
+
 	/* If the irq_work is in use, fall back to report ips. Same
 	 * fallback is used for kernel stack (!user) on a stackmap with
 	 * build_id.
Signature · 1 line
2.54.0
author reviewer
#7 lore raw

PATCH v7 3/3 bpf: Cache build IDs in sleepable stackmap path

In reply to Ihor Solodrai

Stack traces often contain adjacent IPs from the same VMA or from different VMAs backed by the same ELF file. Cache the last successfully parsed build id together with the resolved VMA range and backing file so the sleepable build id path can avoid repeated VMA locking and file parsing in common cases.

---
 kernel/bpf/stackmap.c | 61 ++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 55 insertions(+), 6 deletions(-)
kernel/bpf/stackmap.c +55 -6
index c53cfd9a67cf..77ba03216c09 100644
@@ -246,6 +246,14 @@ static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i
 {
 	struct mm_struct *mm = current->mm;
 	struct stack_map_vma_lock lock = { .mm = mm };
+	struct {
+		struct file *file;
+		const unsigned char *build_id;
+		unsigned long vm_start;
+		unsigned long vm_end;
+		unsigned long vm_pgoff;
+	} cache = {};
+	unsigned long vm_pgoff, vm_start, vm_end;
 	struct vm_area_struct *vma;
 	struct file *file;
 	u64 offset;
@@ -254,6 +262,17 @@ static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i
 	for (u32 i = 0; i < trace_nr; i++) {
 		ip = READ_ONCE(id_offs[i].ip);
 
+		/*
+		 * Range cache fast path: if ip falls within the previously
+		 * resolved VMA range, reuse the cache build_id without
+		 * re-acquiring the VMA lock.
+		 */
+		if (cache.build_id && ip >= cache.vm_start && ip < cache.vm_end) {
+			offset = stack_map_build_id_offset(cache.vm_pgoff, cache.vm_start, ip);
+			stack_map_build_id_set_valid(&id_offs[i], offset, cache.build_id);
+			continue;
+		}
+
 		vma = stack_map_lock_vma(&lock, ip);
 		if (!vma) {
 			stack_map_build_id_set_ip(&id_offs[i]);
@@ -265,17 +284,47 @@ static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i
 			continue;
 		}
 
-		file = get_file(vma->vm_file);
-		offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+		file = vma->vm_file;
+		vm_pgoff = vma->vm_pgoff;
+		vm_start = vma->vm_start;
+		vm_end = vma->vm_end;
+		offset = stack_map_build_id_offset(vm_pgoff, vm_start, ip);
+
+		/*
+		 * Same backing file as previous (e.g. different VMAs
+		 * of the same ELF binary). Reuse the cache build_id.
+		 */
+		if (file == cache.file) {
+			stack_map_unlock_vma(&lock);
+			stack_map_build_id_set_valid(&id_offs[i], offset, cache.build_id);
+			cache.vm_start = vm_start;
+			cache.vm_end = vm_end;
+			cache.vm_pgoff = vm_pgoff;
+			continue;
+		}
+
+		file = get_file(file);
 		stack_map_unlock_vma(&lock);
 
 		/* build_id_parse_file() may block on filesystem reads */
-		if (build_id_parse_file(file, id_offs[i].build_id, NULL))
+		if (build_id_parse_file(file, id_offs[i].build_id, NULL)) {
 			stack_map_build_id_set_ip(&id_offs[i]);
-		else
-			stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
-		fput(file);
+			fput(file);
+			continue;
+		}
+
+		stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
+		if (cache.file)
+			fput(cache.file);
+		cache.file = file;
+		cache.build_id = id_offs[i].build_id;
+		cache.vm_start = vm_start;
+		cache.vm_end = vm_end;
+		cache.vm_pgoff = vm_pgoff;
 	}
+
+	if (cache.file)
+		fput(cache.file);
 }
 
 /*
Signature · 1 line
2.54.0
#8 lore raw

Re: [PATCH bpf-next v7 3/3] bpf: Cache build IDs in sleepable stackmap path

Quoted 39 lines · Ihor Solodrai

Stack traces often contain adjacent IPs from the same VMA or from different VMAs backed by the same ELF file. Cache the last successfully parsed build id together with the resolved VMA range and backing file so the sleepable build id path can avoid repeated VMA locking and file parsing in common cases.

---
 kernel/bpf/stackmap.c | 61 ++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 55 insertions(+), 6 deletions(-)
kernel/bpf/stackmap.c +9 -0
index c53cfd9a67cf..77ba03216c09 100644
@@ -246,6 +246,14 @@ static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i
 {
        struct mm_struct *mm = current->mm;
        struct stack_map_vma_lock lock = { .mm = mm };
+       struct {
+               struct file *file;
+               const unsigned char *build_id;
+               unsigned long vm_start;
+               unsigned long vm_end;
+               unsigned long vm_pgoff;
+       } cache = {};
+       unsigned long vm_pgoff, vm_start, vm_end;
        struct vm_area_struct *vma;
        struct file *file;
        u64 offset;
@@ -254,6 +262,17 @@ static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i
        for (u32 i = 0; i < trace_nr; i++) {
                ip = READ_ONCE(id_offs[i].ip);
 
+               /*

kernel/bpf/stackmap.c static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i · line 275

Quoted 10 lines
kernel/bpf/stackmap.c +10 -0
+                * Range cache fast path: if ip falls within the previously
+                * resolved VMA range, reuse the cache build_id without
+                * re-acquiring the VMA lock.
+                */
+               if (cache.build_id && ip >= cache.vm_start && ip < cache.vm_end) {
+                       offset = stack_map_build_id_offset(cache.vm_pgoff, cache.vm_start, ip);
+                       stack_map_build_id_set_valid(&id_offs[i], offset, cache.build_id);
+                       continue;
+               }
+

Same issue as with patch #1, this is a lost opportunity to avoid unnecessary work if VMA has no matching build ID. We should do negative caching (i.e., have build_id set to NULL, but use vm_start/vm_end for subsequent ip matches) and avoid unnecessary vma lookups. Let's treat it as optimization and address this in a follow up patch.

Quoted 29 lines
kernel/bpf/stackmap.c +20 -2
                vma = stack_map_lock_vma(&lock, ip);
                if (!vma) {
                        stack_map_build_id_set_ip(&id_offs[i]);
@@ -265,17 +284,47 @@ static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i
                        continue;
                }
 
-               file = get_file(vma->vm_file);
-               offset = stack_map_build_id_offset(vma->vm_pgoff, vma->vm_start, ip);
+               file = vma->vm_file;
+               vm_pgoff = vma->vm_pgoff;
+               vm_start = vma->vm_start;
+               vm_end = vma->vm_end;
+               offset = stack_map_build_id_offset(vm_pgoff, vm_start, ip);
+
+               /*
+                * Same backing file as previous (e.g. different VMAs
+                * of the same ELF binary). Reuse the cache build_id.
+                */
+               if (file == cache.file) {
+                       stack_map_unlock_vma(&lock);
+                       stack_map_build_id_set_valid(&id_offs[i], offset, cache.build_id);
+                       cache.vm_start = vm_start;
+                       cache.vm_end = vm_end;
+                       cache.vm_pgoff = vm_pgoff;
+                       continue;
+               }
+
+               file = get_file(file);

kernel/bpf/stackmap.c static void stack_map_get_build_id_offset_sleepable(struct bpf_stack_build_id *i · line 312

Quoted 10 lines
kernel/bpf/stackmap.c +2 -4
                stack_map_unlock_vma(&lock);
 
                /* build_id_parse_file() may block on filesystem reads */
-               if (build_id_parse_file(file, id_offs[i].build_id, NULL))
+               if (build_id_parse_file(file, id_offs[i].build_id, NULL)) {
                        stack_map_build_id_set_ip(&id_offs[i]);
-               else
-                       stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
-               fput(file);
+                       fput(file);

we should probably negative-cache this lack of build ID and update vm_start/vm_end here, in a follow up

Quoted 21 lines
kernel/bpf/stackmap.c +14 -1
+                       continue;
+               }
+
+               stack_map_build_id_set_valid(&id_offs[i], offset, id_offs[i].build_id);
+               if (cache.file)
+                       fput(cache.file);
+               cache.file = file;
+               cache.build_id = id_offs[i].build_id;
+               cache.vm_start = vm_start;
+               cache.vm_end = vm_end;
+               cache.vm_pgoff = vm_pgoff;
        }
+
+       if (cache.file)
+               fput(cache.file);
 }
 
 /*
--

2.54.0

#9 lore raw
In reply to Ihor Solodrai
This series was applied to bpf/bpf-next.git (master) by Andrii Nakryiko <[email protected]>:

Hello:

This series was applied to bpf/bpf-next.git (master) by Andrii Nakryiko <[email protected]>:

Quoted 9 lines · On Mon, 25 May 2026 15:39:45 -0700 you wrote:

The series introduces stack_map_get_build_id_offset_sleepable(), fixing a gap with parsing build_id in sleepable context in stackmap.c

In particular, this fixes a deadlock in stack_map_get_build_id_offset() doing a blocking __kernel_read(), which happens since commit 777a8560fd29 ("lib/buildid: use __kernel_read() for sleepable context").

[...]

Here is the summary with links:
- [bpf-next,v7,1/3] bpf: Factor out stack_map build ID helpers

    https://git.kernel.org/bpf/bpf-next/c/fc99547a8bda
  - [bpf-next,v7,2/3] bpf: Avoid faultable build ID reads under mm locks
    https://git.kernel.org/bpf/bpf-next/c/fad3021faf7b
  - [bpf-next,v7,3/3] bpf: Cache build IDs in sleepable stackmap path
    https://git.kernel.org/bpf/bpf-next/c/5e9099d8ff24

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html