2024-03-07 19:27:22

by Luck, Tony

[permalink] [raw]
Subject: [PATCH v3] x86/mce: Dynamically size space for machine check records

Systems with a large number of CPUs may generate a large
number of machine check records when things go seriously
wrong. But Linux has a fixed buffer that can only capture
a few dozen errors.

Allocate space based on the number of CPUs (with a minimum
value based on the historical fixed buffer that could store
80 records).

Signed-off-by: Tony Luck <[email protected]>
Reviewed-by: Sohil Mehta <[email protected]>
---

Changes since V2; Link: https://lore.kernel.org/all/[email protected]/

Boris: Eliminate "out:" label in mce_gen_pool_create()

Sohil: Added Reviewed-by tag

arch/x86/kernel/cpu/mce/genpool.c | 28 +++++++++++++++++++---------
1 file changed, 19 insertions(+), 9 deletions(-)

diff --git a/arch/x86/kernel/cpu/mce/genpool.c b/arch/x86/kernel/cpu/mce/genpool.c
index fbe8b61c3413..cadf28662a70 100644
--- a/arch/x86/kernel/cpu/mce/genpool.c
+++ b/arch/x86/kernel/cpu/mce/genpool.c
@@ -16,14 +16,14 @@
* used to save error information organized in a lock-less list.
*
* This memory pool is only to be used to save MCE records in MCE context.
- * MCE events are rare, so a fixed size memory pool should be enough. Use
- * 2 pages to save MCE events for now (~80 MCE records at most).
+ * MCE events are rare, so a fixed size memory pool should be enough.
+ * Allocate on a sliding scale based on number of CPUs.
*/
-#define MCE_POOLSZ (2 * PAGE_SIZE)
+#define MCE_MIN_ENTRIES 80
+#define MCE_PER_CPU 2

static struct gen_pool *mce_evt_pool;
static LLIST_HEAD(mce_event_llist);
-static char gen_pool_buf[MCE_POOLSZ];

/*
* Compare the record "t" with each of the records on list "l" to see if
@@ -118,22 +118,32 @@ int mce_gen_pool_add(struct mce *mce)

static int mce_gen_pool_create(void)
{
+ int mce_numrecords, mce_poolsz, order;
struct gen_pool *tmpp;
int ret = -ENOMEM;
+ void *mce_pool;

- tmpp = gen_pool_create(ilog2(sizeof(struct mce_evt_llist)), -1);
+ order = order_base_2(sizeof(struct mce_evt_llist));
+ tmpp = gen_pool_create(order, -1);
if (!tmpp)
- goto out;
+ return ret;

- ret = gen_pool_add(tmpp, (unsigned long)gen_pool_buf, MCE_POOLSZ, -1);
+ mce_numrecords = max(MCE_MIN_ENTRIES, num_possible_cpus() * MCE_PER_CPU);
+ mce_poolsz = mce_numrecords * (1 << order);
+ mce_pool = kmalloc(mce_poolsz, GFP_KERNEL);
+ if (!mce_pool) {
+ gen_pool_destroy(tmpp);
+ return ret;
+ }
+ ret = gen_pool_add(tmpp, (unsigned long)mce_pool, mce_poolsz, -1);
if (ret) {
gen_pool_destroy(tmpp);
- goto out;
+ kfree(mce_pool);
+ return ret;
}

mce_evt_pool = tmpp;

-out:
return ret;
}


base-commit: d206a76d7d2726f3b096037f2079ce0bd3ba329b
--
2.43.0



2024-03-11 05:39:03

by Naik, Avadhut

[permalink] [raw]
Subject: Re: [PATCH v3] x86/mce: Dynamically size space for machine check records



On 3/7/2024 13:27, Tony Luck wrote:
> Systems with a large number of CPUs may generate a large
> number of machine check records when things go seriously
> wrong. But Linux has a fixed buffer that can only capture
> a few dozen errors.
>
> Allocate space based on the number of CPUs (with a minimum
> value based on the historical fixed buffer that could store
> 80 records).
>
> Signed-off-by: Tony Luck <[email protected]>
> Reviewed-by: Sohil Mehta <[email protected]>
> ---
>
> Changes since V2; Link: https://lore.kernel.org/all/[email protected]/
>
> Boris: Eliminate "out:" label in mce_gen_pool_create()
>
> Sohil: Added Reviewed-by tag
>
> arch/x86/kernel/cpu/mce/genpool.c | 28 +++++++++++++++++++---------
> 1 file changed, 19 insertions(+), 9 deletions(-)
>
> diff --git a/arch/x86/kernel/cpu/mce/genpool.c b/arch/x86/kernel/cpu/mce/genpool.c
> index fbe8b61c3413..cadf28662a70 100644
> --- a/arch/x86/kernel/cpu/mce/genpool.c
> +++ b/arch/x86/kernel/cpu/mce/genpool.c
> @@ -16,14 +16,14 @@
> * used to save error information organized in a lock-less list.
> *
> * This memory pool is only to be used to save MCE records in MCE context.
> - * MCE events are rare, so a fixed size memory pool should be enough. Use
> - * 2 pages to save MCE events for now (~80 MCE records at most).
> + * MCE events are rare, so a fixed size memory pool should be enough.
> + * Allocate on a sliding scale based on number of CPUs.
> */
> -#define MCE_POOLSZ (2 * PAGE_SIZE)
> +#define MCE_MIN_ENTRIES 80
> +#define MCE_PER_CPU 2
>
> static struct gen_pool *mce_evt_pool;
> static LLIST_HEAD(mce_event_llist);
> -static char gen_pool_buf[MCE_POOLSZ];
>
> /*
> * Compare the record "t" with each of the records on list "l" to see if
> @@ -118,22 +118,32 @@ int mce_gen_pool_add(struct mce *mce)
>
> static int mce_gen_pool_create(void)
> {
> + int mce_numrecords, mce_poolsz, order;
> struct gen_pool *tmpp;
> int ret = -ENOMEM;
> + void *mce_pool;
>
> - tmpp = gen_pool_create(ilog2(sizeof(struct mce_evt_llist)), -1);
> + order = order_base_2(sizeof(struct mce_evt_llist));
> + tmpp = gen_pool_create(order, -1);
> if (!tmpp)
> - goto out;
> + return ret;
>
> - ret = gen_pool_add(tmpp, (unsigned long)gen_pool_buf, MCE_POOLSZ, -1);
> + mce_numrecords = max(MCE_MIN_ENTRIES, num_possible_cpus() * MCE_PER_CPU);
> + mce_poolsz = mce_numrecords * (1 << order);
> + mce_pool = kmalloc(mce_poolsz, GFP_KERNEL);
> + if (!mce_pool) {
> + gen_pool_destroy(tmpp);
> + return ret;
> + }
> + ret = gen_pool_add(tmpp, (unsigned long)mce_pool, mce_poolsz, -1);
> if (ret) {
> gen_pool_destroy(tmpp);
> - goto out;
> + kfree(mce_pool);
> + return ret;
> }
>
> mce_evt_pool = tmpp;
>
> -out:
> return ret;
> }
>
>
> base-commit: d206a76d7d2726f3b096037f2079ce0bd3ba329b

LGTM!

Reviewed-by: Avadhut Naik <[email protected]>

--
Thanks,
Avadhut Naik

2024-03-26 11:50:12

by Borislav Petkov

[permalink] [raw]
Subject: Re: [PATCH v3] x86/mce: Dynamically size space for machine check records

On Thu, Mar 07, 2024 at 11:27:04AM -0800, Tony Luck wrote:
> - ret = gen_pool_add(tmpp, (unsigned long)gen_pool_buf, MCE_POOLSZ, -1);
> + mce_numrecords = max(MCE_MIN_ENTRIES, num_possible_cpus() * MCE_PER_CPU);
> + mce_poolsz = mce_numrecords * (1 << order);

So, on a big fat machine with 8K CPUs that's, what

8192 * 2 * (1 << 8) = ~4M

buffer?

Well, if you have a fat machine, you get fat buffers too.

--
Regards/Gruss,
Boris.

https://people.kernel.org/tglx/notes-about-netiquette

Subject: [tip: ras/core] x86/mce: Dynamically size space for machine check records

The following commit has been merged into the ras/core branch of tip:

Commit-ID: 108c6494bdf1dfeaefc0a506e2f471aa92fafdd6
Gitweb: https://git.kernel.org/tip/108c6494bdf1dfeaefc0a506e2f471aa92fafdd6
Author: Tony Luck <[email protected]>
AuthorDate: Thu, 07 Mar 2024 11:27:04 -08:00
Committer: Borislav Petkov (AMD) <[email protected]>
CommitterDate: Tue, 26 Mar 2024 12:40:42 +01:00

x86/mce: Dynamically size space for machine check records

Systems with a large number of CPUs may generate a large number of
machine check records when things go seriously wrong. But Linux has
a fixed-size buffer that can only capture a few dozen errors.

Allocate space based on the number of CPUs (with a minimum value based
on the historical fixed buffer that could store 80 records).

[ bp: Rename local var from tmpp to something more telling: gpool. ]

Signed-off-by: Tony Luck <[email protected]>
Signed-off-by: Borislav Petkov (AMD) <[email protected]>
Reviewed-by: Sohil Mehta <[email protected]>
Reviewed-by: Avadhut Naik <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
---
arch/x86/kernel/cpu/mce/genpool.c | 40 ++++++++++++++++++------------
1 file changed, 25 insertions(+), 15 deletions(-)

diff --git a/arch/x86/kernel/cpu/mce/genpool.c b/arch/x86/kernel/cpu/mce/genpool.c
index fbe8b61..4284749 100644
--- a/arch/x86/kernel/cpu/mce/genpool.c
+++ b/arch/x86/kernel/cpu/mce/genpool.c
@@ -16,14 +16,14 @@
* used to save error information organized in a lock-less list.
*
* This memory pool is only to be used to save MCE records in MCE context.
- * MCE events are rare, so a fixed size memory pool should be enough. Use
- * 2 pages to save MCE events for now (~80 MCE records at most).
+ * MCE events are rare, so a fixed size memory pool should be enough.
+ * Allocate on a sliding scale based on number of CPUs.
*/
-#define MCE_POOLSZ (2 * PAGE_SIZE)
+#define MCE_MIN_ENTRIES 80
+#define MCE_PER_CPU 2

static struct gen_pool *mce_evt_pool;
static LLIST_HEAD(mce_event_llist);
-static char gen_pool_buf[MCE_POOLSZ];

/*
* Compare the record "t" with each of the records on list "l" to see if
@@ -118,22 +118,32 @@ int mce_gen_pool_add(struct mce *mce)

static int mce_gen_pool_create(void)
{
- struct gen_pool *tmpp;
+ int mce_numrecords, mce_poolsz, order;
+ struct gen_pool *gpool;
int ret = -ENOMEM;
-
- tmpp = gen_pool_create(ilog2(sizeof(struct mce_evt_llist)), -1);
- if (!tmpp)
- goto out;
-
- ret = gen_pool_add(tmpp, (unsigned long)gen_pool_buf, MCE_POOLSZ, -1);
+ void *mce_pool;
+
+ order = order_base_2(sizeof(struct mce_evt_llist));
+ gpool = gen_pool_create(order, -1);
+ if (!gpool)
+ return ret;
+
+ mce_numrecords = max(MCE_MIN_ENTRIES, num_possible_cpus() * MCE_PER_CPU);
+ mce_poolsz = mce_numrecords * (1 << order);
+ mce_pool = kmalloc(mce_poolsz, GFP_KERNEL);
+ if (!mce_pool) {
+ gen_pool_destroy(gpool);
+ return ret;
+ }
+ ret = gen_pool_add(gpool, (unsigned long)mce_pool, mce_poolsz, -1);
if (ret) {
- gen_pool_destroy(tmpp);
- goto out;
+ gen_pool_destroy(gpool);
+ kfree(mce_pool);
+ return ret;
}

- mce_evt_pool = tmpp;
+ mce_evt_pool = gpool;

-out:
return ret;
}