2010-07-23 17:40:23

by Yinghai Lu

[permalink] [raw]
Subject: [PATCH] acpi: fix apei related table size checking


also check if we can find right action in apei.

Signed-off-by: Yinghai Lu <[email protected]>

---
drivers/acpi/apei/apei-base.c | 5 +++++
drivers/acpi/apei/einj.c | 3 ++-
drivers/acpi/apei/erst.c | 3 ++-
3 files changed, 9 insertions(+), 2 deletions(-)

Index: linux-2.6/drivers/acpi/apei/einj.c
===================================================================
--- linux-2.6.orig/drivers/acpi/apei/einj.c
+++ linux-2.6/drivers/acpi/apei/einj.c
@@ -426,7 +426,8 @@ DEFINE_SIMPLE_ATTRIBUTE(error_inject_fop

static int einj_check_table(struct acpi_table_einj *einj_tab)
{
- if (einj_tab->header_length != sizeof(struct acpi_table_einj))
+ if (einj_tab->header_length !=
+ (sizeof(struct acpi_table_einj) - sizeof(einj_tab->header)))
return -EINVAL;
if (einj_tab->header.length < sizeof(struct acpi_table_einj))
return -EINVAL;
Index: linux-2.6/drivers/acpi/apei/erst.c
===================================================================
--- linux-2.6.orig/drivers/acpi/apei/erst.c
+++ linux-2.6/drivers/acpi/apei/erst.c
@@ -749,7 +749,8 @@ __setup("erst_disable", setup_erst_disab

static int erst_check_table(struct acpi_table_erst *erst_tab)
{
- if (erst_tab->header_length != sizeof(struct acpi_table_erst))
+ if (erst_tab->header_length !=
+ (sizeof(struct acpi_table_erst) - sizeof(erst_tab->header)))
return -EINVAL;
if (erst_tab->header.length < sizeof(struct acpi_table_erst))
return -EINVAL;
Index: linux-2.6/drivers/acpi/apei/apei-base.c
===================================================================
--- linux-2.6.orig/drivers/acpi/apei/apei-base.c
+++ linux-2.6/drivers/acpi/apei/apei-base.c
@@ -163,6 +163,7 @@ int apei_exec_run(struct apei_exec_conte
u32 i, ip;
struct acpi_whea_header *entry;
apei_exec_ins_func_t run;
+ bool found_action = false;

ctx->ip = 0;

@@ -178,6 +179,7 @@ rewind:
entry = &ctx->action_table[i];
if (entry->action != action)
continue;
+ found_action = true;
if (ip == ctx->ip) {
if (entry->instruction >= ctx->instructions ||
!ctx->ins_table[entry->instruction].run) {
@@ -198,6 +200,9 @@ rewind:
goto rewind;
}

+ if (!found_action)
+ return -ENODEV;
+
return 0;
}
EXPORT_SYMBOL_GPL(apei_exec_run);


2010-07-26 04:23:18

by Bjorn Helgaas

[permalink] [raw]
Subject: Re: [PATCH] acpi: fix apei related table size checking

On Friday, July 23, 2010 11:39:17 am Yinghai Lu wrote:
>
> also check if we can find right action in apei.
>
> Signed-off-by: Yinghai Lu <[email protected]>

This should be split into two patches that each fix one thing.

> ---
> drivers/acpi/apei/apei-base.c | 5 +++++
> drivers/acpi/apei/einj.c | 3 ++-
> drivers/acpi/apei/erst.c | 3 ++-
> 3 files changed, 9 insertions(+), 2 deletions(-)
>
> Index: linux-2.6/drivers/acpi/apei/einj.c
> ===================================================================
> --- linux-2.6.orig/drivers/acpi/apei/einj.c
> +++ linux-2.6/drivers/acpi/apei/einj.c
> @@ -426,7 +426,8 @@ DEFINE_SIMPLE_ATTRIBUTE(error_inject_fop
>
> static int einj_check_table(struct acpi_table_einj *einj_tab)
> {
> - if (einj_tab->header_length != sizeof(struct acpi_table_einj))
> + if (einj_tab->header_length !=
> + (sizeof(struct acpi_table_einj) - sizeof(einj_tab->header)))
> return -EINVAL;
> if (einj_tab->header.length < sizeof(struct acpi_table_einj))
> return -EINVAL;
> Index: linux-2.6/drivers/acpi/apei/erst.c
> ===================================================================
> --- linux-2.6.orig/drivers/acpi/apei/erst.c
> +++ linux-2.6/drivers/acpi/apei/erst.c
> @@ -749,7 +749,8 @@ __setup("erst_disable", setup_erst_disab
>
> static int erst_check_table(struct acpi_table_erst *erst_tab)
> {
> - if (erst_tab->header_length != sizeof(struct acpi_table_erst))
> + if (erst_tab->header_length !=
> + (sizeof(struct acpi_table_erst) - sizeof(erst_tab->header)))
> return -EINVAL;
> if (erst_tab->header.length < sizeof(struct acpi_table_erst))
> return -EINVAL;
> Index: linux-2.6/drivers/acpi/apei/apei-base.c
> ===================================================================
> --- linux-2.6.orig/drivers/acpi/apei/apei-base.c
> +++ linux-2.6/drivers/acpi/apei/apei-base.c
> @@ -163,6 +163,7 @@ int apei_exec_run(struct apei_exec_conte
> u32 i, ip;
> struct acpi_whea_header *entry;
> apei_exec_ins_func_t run;
> + bool found_action = false;
>
> ctx->ip = 0;
>
> @@ -178,6 +179,7 @@ rewind:
> entry = &ctx->action_table[i];
> if (entry->action != action)
> continue;
> + found_action = true;
> if (ip == ctx->ip) {
> if (entry->instruction >= ctx->instructions ||
> !ctx->ins_table[entry->instruction].run) {
> @@ -198,6 +200,9 @@ rewind:
> goto rewind;
> }
>
> + if (!found_action)
> + return -ENODEV;
> +
> return 0;
> }
> EXPORT_SYMBOL_GPL(apei_exec_run);
> --
> To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>

2010-07-26 07:13:23

by Andi Kleen

[permalink] [raw]
Subject: Re: [PATCH] acpi: fix apei related table size checking

On 7/26/2010 6:23 AM, Bjorn Helgaas wrote:
> On Friday, July 23, 2010 11:39:17 am Yinghai Lu wrote:
>> also check if we can find right action in apei.
>>
>> Signed-off-by: Yinghai Lu<[email protected]>
> This should be split into two patches that each fix one thing.

... and the description actually describe what it is supposed to fix.

-Andi

2010-07-26 07:41:45

by Huang, Ying

[permalink] [raw]
Subject: Re: [PATCH] acpi: fix apei related table size checking

On Sat, 2010-07-24 at 01:39 +0800, Yinghai Lu wrote:
> also check if we can find right action in apei.
>
> Signed-off-by: Yinghai Lu <[email protected]>
>
> ---
> drivers/acpi/apei/apei-base.c | 5 +++++
> drivers/acpi/apei/einj.c | 3 ++-
> drivers/acpi/apei/erst.c | 3 ++-
> 3 files changed, 9 insertions(+), 2 deletions(-)
>
> Index: linux-2.6/drivers/acpi/apei/einj.c
> ===================================================================
> --- linux-2.6.orig/drivers/acpi/apei/einj.c
> +++ linux-2.6/drivers/acpi/apei/einj.c
> @@ -426,7 +426,8 @@ DEFINE_SIMPLE_ATTRIBUTE(error_inject_fop
>
> static int einj_check_table(struct acpi_table_einj *einj_tab)
> {
> - if (einj_tab->header_length != sizeof(struct acpi_table_einj))
> + if (einj_tab->header_length !=
> + (sizeof(struct acpi_table_einj) - sizeof(einj_tab->header)))
> return -EINVAL;

It seems that the header_length field is not set properly by BIOS on my
testing machine which I used to develop the original code. Will try to
contact BIOS guys to make sure.

> Index: linux-2.6/drivers/acpi/apei/apei-base.c
> ===================================================================
> --- linux-2.6.orig/drivers/acpi/apei/apei-base.c
> +++ linux-2.6/drivers/acpi/apei/apei-base.c
> @@ -163,6 +163,7 @@ int apei_exec_run(struct apei_exec_conte
> u32 i, ip;
> struct acpi_whea_header *entry;
> apei_exec_ins_func_t run;
> + bool found_action = false;
>
> ctx->ip = 0;
>
> @@ -178,6 +179,7 @@ rewind:
> entry = &ctx->action_table[i];
> if (entry->action != action)
> continue;
> + found_action = true;
> if (ip == ctx->ip) {
> if (entry->instruction >= ctx->instructions ||
> !ctx->ins_table[entry->instruction].run) {
> @@ -198,6 +200,9 @@ rewind:
> goto rewind;
> }
>
> + if (!found_action)
> + return -ENODEV;

-ENOENT is better here?

Best Regards,
Huang Ying