Skip to contentSkip to conversation

ring-buffer: Do not check si_mem_available() before deferred init completes

PATCH Superseded Amit Machhiwal 3 participants 5 messages Oct 23 – Nov 18

Subsystems (4)

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

PATCH v4 1/1 ring-buffer: Do not check si_mem_available() before deferred init completes

During early boot (early_trace_init() called from start_kernel()), tracing allocates initial ring buffers (temp_buffer, array_buffer, and snapshot_buffer) for CPU 0. Later, as secondary CPUs come online during smp_init(), each CPU's ring buffer is allocated via trace_rb_cpu_prepare().

Before attempting page allocation, __rb_allocate_pages() performs a heuristic check using si_mem_available() to return early with -ENOMEM if memory appears insufficient.

However, on kernels built with CONFIG_DEFERRED_STRUCT_PAGE_INIT=y, NR_FREE_PAGES only reflects the initial non-deferred pool across the entire period from early_trace_init() through smp_init(). defer_init() leaves only a single section per node (e.g. 16 MiB with 64 KB pages) initialized up-front, and on kernels with large static binary footprints (debug configurations enabling PAGE_OWNER, DEBUG_PAGEALLOC, KFENCE, or SLUB_DEBUG), early core initialisations (SLUB caches, vmalloc, static ftrace records) consume virtually all managed pages in this pool.

si_mem_available() therefore returns a false negative (often 0-1 pages) across both SYSTEM_BOOTING (for the initial CPU 0 buffers) and SYSTEM_SCHEDULING (for secondary CPU buffers via trace_rb_cpu_prepare()), causing __rb_allocate_pages() to prematurely abort with -ENOMEM.

This failure is false: the buddy allocator is designed to replenish ZONE_NORMAL on demand via deferred_grow_zone() when an actual allocation is attempted. Checking si_mem_available() before calling alloc_pages_node() short-circuits this on-demand growth entirely.

deferred_grow_zone() is only permanently disabled after page_alloc_init_late() completes (it spawns per-node kthreads and blocks until done), which happens inside kernel_init_freeable() — before system_state is set to SYSTEM_FREEING_INITMEM. Skip the si_mem_available() check until SYSTEM_FREEING_INITMEM, at which point all deferred pages are initialized and si_mem_available() is accurate.

Use IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) to set i = LONG_MAX during this window, unconditionally bypassing the threshold check. On kernels without deferred init IS_ENABLED() compiles the branch away and the original check is preserved unchanged.

---
Changes since v3:
- Replace #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT / #else / #endif with IS_ENABLED() + i = LONG_MAX to avoid duplicating the threshold check. (Vincent Donnefort)
- Trim the in-code comment to one sentence. (Vincent Donnefort)

Changes since v2:
- Rebase on top of upstream ring_buffer.c changes from Vincent Donnefort; update threshold to i < (nr_pages << subbuf_order). (Michal Suchánek)

Changes since v1:
- Extend the guard to cover SYSTEM_SCHEDULING as well: smp_init() calls trace_rb_cpu_prepare() -> __rb_allocate_pages() before page_alloc_init_late() completes, so the false negative window spans both SYSTEM_BOOTING and SYSTEM_SCHEDULING. Use system_state >= SYSTEM_FREEING_INITMEM as the correct boundary. (Sashiko, Vincent Donnefort)
- Wrap in #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT to preserve the original check unchanged on non-deferred kernels. (Vincent Donnefort)
- Drop Tested-by: guard condition changed, requesting fresh test.

 kernel/trace/ring_buffer.c | 8 ++++++++
 1 file changed, 8 insertions(+)
kernel/trace/ring_buffer.c +8 -0
index 9c4e40847e43..09c02e838c4d 100644
@@ -2463,8 +2463,16 @@ static int __rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer,
 		 * memory. It may not be accurate. But we don't care, we just want
 		 * to prevent doing any allocation when it is obvious that it is
 		 * not going to succeed.
+		 *
+		 * With CONFIG_DEFERRED_STRUCT_PAGE_INIT, si_mem_available()
+		 * will reflect the actual available memory only after
+		 * SYSTEM_FREEING_INITMEM.
 		 */
 		i = si_mem_available();
+		if (IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) &&
+		    system_state < SYSTEM_FREEING_INITMEM)
+			i = LONG_MAX;
+
 		if (i < (nr_pages << subbuf_order))
 			return -ENOMEM;
 

base-commit: 734bef548a96e8774d6eef5dfa2026c6edc4e80b

Signature · 1 line
2.54.0 (Apple Git-157)
maintainer Reviewed-by on v4 1/1
#1 lore raw
Quoted 35 lines · Amit Machhiwal

During early boot (early_trace_init() called from start_kernel()), tracing allocates initial ring buffers (temp_buffer, array_buffer, and snapshot_buffer) for CPU 0. Later, as secondary CPUs come online during smp_init(), each CPU's ring buffer is allocated via trace_rb_cpu_prepare().

Before attempting page allocation, __rb_allocate_pages() performs a heuristic check using si_mem_available() to return early with -ENOMEM if memory appears insufficient.

However, on kernels built with CONFIG_DEFERRED_STRUCT_PAGE_INIT=y, NR_FREE_PAGES only reflects the initial non-deferred pool across the entire period from early_trace_init() through smp_init(). defer_init() leaves only a single section per node (e.g. 16 MiB with 64 KB pages) initialized up-front, and on kernels with large static binary footprints (debug configurations enabling PAGE_OWNER, DEBUG_PAGEALLOC, KFENCE, or SLUB_DEBUG), early core initialisations (SLUB caches, vmalloc, static ftrace records) consume virtually all managed pages in this pool.

si_mem_available() therefore returns a false negative (often 0-1 pages) across both SYSTEM_BOOTING (for the initial CPU 0 buffers) and SYSTEM_SCHEDULING (for secondary CPU buffers via trace_rb_cpu_prepare()), causing __rb_allocate_pages() to prematurely abort with -ENOMEM.

This failure is false: the buddy allocator is designed to replenish ZONE_NORMAL on demand via deferred_grow_zone() when an actual allocation is attempted. Checking si_mem_available() before calling alloc_pages_node() short-circuits this on-demand growth entirely.

deferred_grow_zone() is only permanently disabled after page_alloc_init_late() completes (it spawns per-node kthreads and blocks until done), which happens inside kernel_init_freeable() — before system_state is set to SYSTEM_FREEING_INITMEM. Skip the si_mem_available() check until SYSTEM_FREEING_INITMEM, at which point all deferred pages are initialized and si_mem_available() is accurate.

Quoted 4 lines

Use IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) to set i = LONG_MAX during this window, unconditionally bypassing the threshold check. On kernels without deferred init IS_ENABLED() compiles the branch away and the original check is preserved unchanged.

Commit messages are not LLMs' strong suit.

Otherwise.

Quoted 57 lines

---
Changes since v3:
- Replace #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT / #else / #endif with IS_ENABLED() + i = LONG_MAX to avoid duplicating the threshold check. (Vincent Donnefort)
- Trim the in-code comment to one sentence. (Vincent Donnefort)

Changes since v2:
- Rebase on top of upstream ring_buffer.c changes from Vincent Donnefort; update threshold to i < (nr_pages << subbuf_order). (Michal Suchánek)

Changes since v1:
- Extend the guard to cover SYSTEM_SCHEDULING as well: smp_init() calls trace_rb_cpu_prepare() -> __rb_allocate_pages() before page_alloc_init_late() completes, so the false negative window spans both SYSTEM_BOOTING and SYSTEM_SCHEDULING. Use system_state >= SYSTEM_FREEING_INITMEM as the correct boundary. (Sashiko, Vincent Donnefort)
- Wrap in #ifdef CONFIG_DEFERRED_STRUCT_PAGE_INIT to preserve the original check unchanged on non-deferred kernels. (Vincent Donnefort)
- Drop Tested-by: guard condition changed, requesting fresh test.

 kernel/trace/ring_buffer.c | 8 ++++++++
 1 file changed, 8 insertions(+)
kernel/trace/ring_buffer.c +8 -0
index 9c4e40847e43..09c02e838c4d 100644
@@ -2463,8 +2463,16 @@ static int __rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer,
 		 * memory. It may not be accurate. But we don't care, we just want
 		 * to prevent doing any allocation when it is obvious that it is
 		 * not going to succeed.
+		 *
+		 * With CONFIG_DEFERRED_STRUCT_PAGE_INIT, si_mem_available()
+		 * will reflect the actual available memory only after
+		 * SYSTEM_FREEING_INITMEM.
 		 */
 		i = si_mem_available();
+		if (IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) &&
+		    system_state < SYSTEM_FREEING_INITMEM)
+			i = LONG_MAX;
+
 		if (i < (nr_pages << subbuf_order))
 			return -ENOMEM;
 

base-commit: 734bef548a96e8774d6eef5dfa2026c6edc4e80b

Signature · 1 line
2.54.0 (Apple Git-157)
#2 lore raw
In reply to Amit Machhiwal

Hi Steven,

Gentle ping on this patch. This has Vincent's R-b and fixes a boot-time ring-buffer allocation failure on DEFERRED_STRUCT_PAGE_INIT kernels. Is there anything else needed from my side? I'm happy to trim the commit message or send a v5 if that would help.

Thanks,
Amit

maintainer
#3 lore raw
In reply to Amit Machhiwal
Quoted 35 lines · On Fri, 23 Oct 2026 20:24:59 +0530 Amit Machhiwal <[email protected]>…

During early boot (early_trace_init() called from start_kernel()), tracing allocates initial ring buffers (temp_buffer, array_buffer, and snapshot_buffer) for CPU 0. Later, as secondary CPUs come online during smp_init(), each CPU's ring buffer is allocated via trace_rb_cpu_prepare().

Before attempting page allocation, __rb_allocate_pages() performs a heuristic check using si_mem_available() to return early with -ENOMEM if memory appears insufficient.

However, on kernels built with CONFIG_DEFERRED_STRUCT_PAGE_INIT=y, NR_FREE_PAGES only reflects the initial non-deferred pool across the entire period from early_trace_init() through smp_init(). defer_init() leaves only a single section per node (e.g. 16 MiB with 64 KB pages) initialized up-front, and on kernels with large static binary footprints (debug configurations enabling PAGE_OWNER, DEBUG_PAGEALLOC, KFENCE, or SLUB_DEBUG), early core initialisations (SLUB caches, vmalloc, static ftrace records) consume virtually all managed pages in this pool.

si_mem_available() therefore returns a false negative (often 0-1 pages) across both SYSTEM_BOOTING (for the initial CPU 0 buffers) and SYSTEM_SCHEDULING (for secondary CPU buffers via trace_rb_cpu_prepare()), causing __rb_allocate_pages() to prematurely abort with -ENOMEM.

This failure is false: the buddy allocator is designed to replenish ZONE_NORMAL on demand via deferred_grow_zone() when an actual allocation is attempted. Checking si_mem_available() before calling alloc_pages_node() short-circuits this on-demand growth entirely.

deferred_grow_zone() is only permanently disabled after page_alloc_init_late() completes (it spawns per-node kthreads and blocks until done), which happens inside kernel_init_freeable() — before system_state is set to SYSTEM_FREEING_INITMEM. Skip the si_mem_available() check until SYSTEM_FREEING_INITMEM, at which point all deferred pages are initialized and si_mem_available() is accurate.

Quoted 4 lines

Use IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) to set i = LONG_MAX during this window, unconditionally bypassing the threshold check. On kernels without deferred init IS_ENABLED() compiles the branch away and the original check is preserved unchanged.

As Vincent mentioned, the above is just word salad. Please make it human readable. That is, state what the actual issue is and how it's fixed and remove all the AI fluff.

Also note, just because we no longer require an "Assisted-by" tag, it should still be mentioned how the patch was created. I want to know.

Quoted 16 lines

---

kernel/trace/ring_buffer.c +0 -0
index 9c4e40847e43..09c02e838c4d 100644
@@ -2463,8 +2463,16 @@ static int __rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer,
 		 * memory. It may not be accurate. But we don't care, we just want
 		 * to prevent doing any allocation when it is obvious that it is

kernel/trace/ring_buffer.c static int __rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer, · line 2474

Quoted 10 lines
kernel/trace/ring_buffer.c +7 -0
 		 * not going to succeed.
+		 *
+		 * With CONFIG_DEFERRED_STRUCT_PAGE_INIT, si_mem_available()
+		 * will reflect the actual available memory only after
+		 * SYSTEM_FREEING_INITMEM.
 		 */
 		i = si_mem_available();
+		if (IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) &&
+		    system_state < SYSTEM_FREEING_INITMEM)
+			i = LONG_MAX;

Make the above an if/else statement.

		if (IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) &&
		    unlikely(system_state < SYSTEM_FREEING_INITMEM))
			i = LONG_MAX;
		else
	  		i = si_mem_available();

-- Steve

Quoted 6 lines
kernel/trace/ring_buffer.c +1 -0
+
 		if (i < (nr_pages << subbuf_order))
 			return -ENOMEM;
 

base-commit: 734bef548a96e8774d6eef5dfa2026c6edc4e80b

#4 lore raw
Quoted 43 lines · Steven Rostedt
Quoted 35 lines · On Fri, 23 Oct 2026 20:24:59 +0530 Amit Machhiwal <[email protected]>…

During early boot (early_trace_init() called from start_kernel()), tracing allocates initial ring buffers (temp_buffer, array_buffer, and snapshot_buffer) for CPU 0. Later, as secondary CPUs come online during smp_init(), each CPU's ring buffer is allocated via trace_rb_cpu_prepare().

Before attempting page allocation, __rb_allocate_pages() performs a heuristic check using si_mem_available() to return early with -ENOMEM if memory appears insufficient.

However, on kernels built with CONFIG_DEFERRED_STRUCT_PAGE_INIT=y, NR_FREE_PAGES only reflects the initial non-deferred pool across the entire period from early_trace_init() through smp_init(). defer_init() leaves only a single section per node (e.g. 16 MiB with 64 KB pages) initialized up-front, and on kernels with large static binary footprints (debug configurations enabling PAGE_OWNER, DEBUG_PAGEALLOC, KFENCE, or SLUB_DEBUG), early core initialisations (SLUB caches, vmalloc, static ftrace records) consume virtually all managed pages in this pool.

si_mem_available() therefore returns a false negative (often 0-1 pages) across both SYSTEM_BOOTING (for the initial CPU 0 buffers) and SYSTEM_SCHEDULING (for secondary CPU buffers via trace_rb_cpu_prepare()), causing __rb_allocate_pages() to prematurely abort with -ENOMEM.

This failure is false: the buddy allocator is designed to replenish ZONE_NORMAL on demand via deferred_grow_zone() when an actual allocation is attempted. Checking si_mem_available() before calling alloc_pages_node() short-circuits this on-demand growth entirely.

deferred_grow_zone() is only permanently disabled after page_alloc_init_late() completes (it spawns per-node kthreads and blocks until done), which happens inside kernel_init_freeable() — before system_state is set to SYSTEM_FREEING_INITMEM. Skip the si_mem_available() check until SYSTEM_FREEING_INITMEM, at which point all deferred pages are initialized and si_mem_available() is accurate.

Quoted 4 lines

Use IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) to set i = LONG_MAX during this window, unconditionally bypassing the threshold check. On kernels without deferred init IS_ENABLED() compiles the branch away and the original check is preserved unchanged.

Quoted 6 lines

As Vincent mentioned, the above is just word salad. Please make it human readable. That is, state what the actual issue is and how it's fixed and remove all the AI fluff.

Also note, just because we no longer require an "Assisted-by" tag, it should still be mentioned how the patch was created. I want to know.

Hi Steven,

To answer how this was created: I investigated the issue, carried out the RCA and debug runs on a Power11 KVM setup, and authored the patch logic. When writing the commit message, in a (clearly failed) attempt to explain the details thoroughly, I ended up introducing the wordiness and fluff. Later I did use an LLM for a quick rewording and spell checking, which only made it worse I guess.

For v5, I have completely rewritten the commit message myself to be concise and to the point: stating the actual issue (deferred page init leaves NR_FREE_PAGES near zero, making si_mem_available() under-report), the spurious -ENOMEM failure during early_trace_init() and smp_init(), and the fix mechanism.

<snip>

Quoted 7 lines

kernel/trace/ring_buffer.c static int __rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer, · line 2474

Quoted 4 lines
kernel/trace/ring_buffer.c +3 -0
 		i = si_mem_available();
+		if (IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) &&
+		    system_state < SYSTEM_FREEING_INITMEM)
+			i = LONG_MAX;

Make the above an if/else statement.

Quoted 5 lines
		if (IS_ENABLED(CONFIG_DEFERRED_STRUCT_PAGE_INIT) &&
		    unlikely(system_state < SYSTEM_FREEING_INITMEM))
			i = LONG_MAX;
		else
	  		i = si_mem_available();

Sure, this makes more sense and avoids calling si_mem_available() when it is not reliable anyway. I have incorporated this in v5.

Thanks,
Amit