[PATCH v2] x86/mce: Always save severity in machine_check_poll

[PATCH v2] x86/mce: Always save severity in machine_check_poll

[Yazen Ghannam]

From: Yazen Ghannam <yazen.ghannam@amd.com>

The severity gives a hint as to how to handle the error. The notifier
blocks can then use the severity to decide on an action. It's not necessary
for machine_check_poll() to filter errors for the notifier chain, since
each block will check its own set of conditions before handling an error.
Also, there isn't any urgency for machine_check_poll() to make decisions
based on severity like in do_machine_check().

If we can assume that a severity is set then we can use them in more
notifier blocks. For example, the CEC block can check for a "KEEP" severity
rather than checking bits in the status. This isn't possible now since the
severity is not set except for "DEFFRRED/UCNA" errors with a valid address.

Save the severity since we have it, and let the notifier blocks decide if
they want to do anything.

Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
---
Link:
https://lkml.kernel.org/r/1497286446-59533-1-git-send-email-Yazen.Ghannam@amd.com

v1->v2:
* Expand commit message.

 arch/x86/kernel/cpu/mcheck/mce.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index b58b77808ce4..6dde0497efc7 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -673,7 +673,6 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 {
 	bool error_seen = false;
 	struct mce m;
-	int severity;
 	int i;
 
 	this_cpu_inc(mce_poll_count);
@@ -710,11 +709,7 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 
 		mce_read_aux(&m, i);
 
-		severity = mce_severity(&m, mca_cfg.tolerant, NULL, false);
-
-		if (severity == MCE_DEFERRED_SEVERITY && mce_is_memory_error(&m))
-			if (m.status & MCI_STATUS_ADDRV)
-				m.severity = severity;
+		m.severity = mce_severity(&m, mca_cfg.tolerant, NULL, false);
 
 		/*
 		 * Don't get the IP here because it's unlikely to
-- 
2.7.4

Re: [PATCH v2] x86/mce: Always save severity in machine_check_poll

[Borislav Petkov]

On Wed, Jun 21, 2017 at 02:46:42PM -0500, Yazen Ghannam wrote:
> From: Yazen Ghannam <yazen.ghannam@amd.com>
> 
> The severity gives a hint as to how to handle the error. The notifier
> blocks can then use the severity to decide on an action. It's not necessary
> for machine_check_poll() to filter errors for the notifier chain, since
> each block will check its own set of conditions before handling an error.
> Also, there isn't any urgency for machine_check_poll() to make decisions
> based on severity like in do_machine_check().
> 
> If we can assume that a severity is set then we can use them in more
> notifier blocks. For example, the CEC block can check for a "KEEP" severity

Hmm, I'm not sure about this. If we did this, we would have to *make*
CECCs be always KEEP severity. And they are now but doing that would
make that mapping CECC->KEEP imprecise and unnecessary but required....

Yeah, we can talk about that when we have to cross that bridge.

So I've left the example but made the above s/can/could/ :)

> rather than checking bits in the status. This isn't possible now since the
> severity is not set except for "DEFFRRED/UCNA" errors with a valid address.
> 
> Save the severity since we have it, and let the notifier blocks decide if
> they want to do anything.
> 
> Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
> ---
> Link:
> https://lkml.kernel.org/r/1497286446-59533-1-git-send-email-Yazen.Ghannam@amd.com
> 
> v1->v2:
> * Expand commit message.
> 
>  arch/x86/kernel/cpu/mcheck/mce.c | 7 +------
>  1 file changed, 1 insertion(+), 6 deletions(-)

Applied, thanks.

-- 
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.

[tip:ras/core] x86/mce: Always save severity in machine_check_poll()

[tip-bot for Yazen Ghannam]

Commit-ID:  e2de64ec52659870b4fdef5bf08f265ce5fe1ccc
Gitweb:     http://git.kernel.org/tip/e2de64ec52659870b4fdef5bf08f265ce5fe1ccc
Author:     Yazen Ghannam <yazen.ghannam@amd.com>
AuthorDate: Mon, 26 Jun 2017 14:35:31 +0200
Committer:  Thomas Gleixner <tglx@linutronix.de>
CommitDate: Mon, 26 Jun 2017 15:58:56 +0200

x86/mce: Always save severity in machine_check_poll()

The MCE severity gives a hint as to how to handle the error. The
notifier blocks can then use the severity to decide on an action.
It's not necessary for machine_check_poll() to filter errors for
the notifier chain, since each block will check its own set of
conditions before handling an error.

Also, there isn't any urgency for machine_check_poll() to make decisions
based on severity like in do_machine_check().

If we can assume that a severity is set then we can use it in more
notifier blocks. For example, the CEC block could check for a "KEEP"
severity rather than checking bits in the status. This isn't possible
now since the severity is not set except for "DEFFRRED/UCNA" errors with
a valid address.

Save the severity since we have it, and let the notifier blocks decide
if they want to do anything.

Signed-off-by: Yazen Ghannam <yazen.ghannam@amd.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1498074402-98633-1-git-send-email-Yazen.Ghannam@amd.com

---
 arch/x86/kernel/cpu/mcheck/mce.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/arch/x86/kernel/cpu/mcheck/mce.c b/arch/x86/kernel/cpu/mcheck/mce.c
index b58b778..6dde049 100644
--- a/arch/x86/kernel/cpu/mcheck/mce.c
+++ b/arch/x86/kernel/cpu/mcheck/mce.c
@@ -673,7 +673,6 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 {
 	bool error_seen = false;
 	struct mce m;
-	int severity;
 	int i;
 
 	this_cpu_inc(mce_poll_count);
@@ -710,11 +709,7 @@ bool machine_check_poll(enum mcp_flags flags, mce_banks_t *b)
 
 		mce_read_aux(&m, i);
 
-		severity = mce_severity(&m, mca_cfg.tolerant, NULL, false);
-
-		if (severity == MCE_DEFERRED_SEVERITY && mce_is_memory_error(&m))
-			if (m.status & MCI_STATUS_ADDRV)
-				m.severity = severity;
+		m.severity = mce_severity(&m, mca_cfg.tolerant, NULL, false);
 
 		/*
 		 * Don't get the IP here because it's unlikely to