Prev: tracing: Add alignment to syscall metadata declarations
Next: [PATCH] arm: Build with -fno-dwarf2-cfi-asm
From: Carlos Corbacho on 9 Jul 2010 17:00
On Friday 09 July 2010 04:51:15 Axel Lin wrote:
> When acpi_evaluate_object() is passed ACPI_ALLOCATE_BUFFER,
> the caller must kfree the returned buffer if AE_OK is returned.
>
> Call Trace:
> wmab_execute
> -> wmi_evaluate_method
> -> acpi_evaluate_object
>
> Thus if callers of wmab_execute() pass ACPI_ALLOCATE_BUFFER,
> the return buffer must be kfreed if wmab_execute return AE_OK.
>
> Signed-off-by: Axel Lin <axel.lin(a)gmail.com>

Acked-by: Carlos Corbacho <carlos(a)strangeworlds.co.uk>
--
E-Mail: carlos(a)strangeworlds.co.uk
Web: strangeworlds.co.uk
GPG Key ID: 0x23EE722D
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Axel Lin on 12 Jul 2010 22:20
2010/7/13 Andrew Morton <akpm(a)linux-foundation.org>:
> On Fri, 09 Jul 2010 12:37:36 +0800 Axel Lin <axel.lin(a)gmail.com> wrote:
>
>> When acpi_evaluate_object() is passed ACPI_ALLOCATE_BUFFER,
>> the caller must kfree the returned buffer if AE_OK is returned.
>>
>> Call Trace:
>> wmab_execute
>> � -> wmi_evaluate_method
>> � � -> acpi_evaluate_object
>>
>> Thus if callers of wmab_execute() pass ACPI_ALLOCATE_BUFFER,
>> the return buffer must be kfreed if wmab_execute return AE_OK.
>>
>> Signed-off-by: Axel Lin <axel.lin(a)gmail.com>
>> ---
>> �drivers/platform/x86/acer-wmi.c | � 11 ++++++++++-
>> �1 files changed, 10 insertions(+), 1 deletions(-)
>>
>> diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
>> index 1ea6c43..3f44446 100644
>> --- a/drivers/platform/x86/acer-wmi.c
>> +++ b/drivers/platform/x86/acer-wmi.c
>> @@ -555,6 +555,7 @@ static acpi_status AMW0_find_mailled(void)
>> � � � obj->buffer.length == sizeof(struct wmab_ret)) {
>> � � � � � � � ret = *((struct wmab_ret *) obj->buffer.pointer);
>> � � � } else {
>> + � � � � � � kfree(out.pointer);
>> � � � � � � � return AE_ERROR;
>> � � � }
>>
>> @@ -598,6 +599,7 @@ static acpi_status AMW0_set_capabilities(void)
>> � � � obj->buffer.length == sizeof(struct wmab_ret)) {
>> � � � � � � � ret = *((struct wmab_ret *) obj->buffer.pointer);
>> � � � } else {
>> + � � � � � � kfree(out.pointer);
>> � � � � � � � return AE_ERROR;
>> � � � }
>>
>> @@ -607,15 +609,22 @@ static acpi_status AMW0_set_capabilities(void)
>> � � � args.ebx = 2 << 8;
>> � � � args.ebx |= ACER_AMW0_BLUETOOTH_MASK;
>>
>> + � � /*
>> + � � �* It's ok to use existing buffer for next wmab_execute call.
>> + � � �* But we need to kfree(out.pointer) if next wmab_execute fail.
>> + � � �*/
>> � � � status = wmab_execute(&args, &out);
>> - � � if (ACPI_FAILURE(status))
>> + � � if (ACPI_FAILURE(status)) {
>> + � � � � � � kfree(out.pointer);
>> � � � � � � � return status;
>> + � � }
>>
>> � � � obj = (union acpi_object *) out.pointer;
>> � � � if (obj && obj->type == ACPI_TYPE_BUFFER
>> � � � && obj->buffer.length == sizeof(struct wmab_ret)) {
>> � � � � � � � ret = *((struct wmab_ret *) obj->buffer.pointer);
>> � � � } else {
>> + � � � � � � kfree(out.pointer);
>> � � � � � � � return AE_ERROR;
>> � � � }
>
> I think it's best to remove the multiple-return-points approach in
> favour of the "goto single return point which undoes things" approach.
> that way we reduce the risk of later introducing more leaks.
>
> Please review:
>
> From: Andrew Morton <akpm(a)linux-foundation.org>
>
> avoid multiple return points
> remove unneeded cast
> remove unneeded initialisation of `status'
>
> Cc: Alan Jenkins <alan-jenkins(a)tuffmail.co.uk>
> Cc: Axel Lin <axel.lin(a)gmail.com>
> Cc: Carlos Corbacho <carlos(a)strangeworlds.co.uk>
> Cc: Matthew Garrett <mjg(a)redhat.com>
> Cc: Thomas Renninger <trenn(a)suse.de>
> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org>

I'm not the author, but since I'm in the CC list and it looks good to me.
Acked-by: Axel Lin <axel.lin(a)gmail.com>

> ---
>
> �drivers/platform/x86/acer-wmi.c | � 25 ++++++++++++-------------
> �1 file changed, 12 insertions(+), 13 deletions(-)
>
> diff -puN drivers/platform/x86/acer-wmi.c~acer-wmi-fix-memory-leaks-in-wmab_execute-error-path-v2-fix drivers/platform/x86/acer-wmi.c
> --- a/drivers/platform/x86/acer-wmi.c~acer-wmi-fix-memory-leaks-in-wmab_execute-error-path-v2-fix
> +++ a/drivers/platform/x86/acer-wmi.c
> @@ -571,7 +571,7 @@ static acpi_status AMW0_set_capabilities
> �{
> � � � �struct wmab_args args;
> � � � �struct wmab_ret ret;
> - � � � acpi_status status = AE_OK;
> + � � � acpi_status status;
> � � � �struct acpi_buffer out = { ACPI_ALLOCATE_BUFFER, NULL };
> � � � �union acpi_object *obj;
>
> @@ -594,13 +594,13 @@ static acpi_status AMW0_set_capabilities
> � � � �if (ACPI_FAILURE(status))
> � � � � � � � �return status;
>
> - � � � obj = (union acpi_object *) out.pointer;
> + � � � obj = out.pointer;
> � � � �if (obj && obj->type == ACPI_TYPE_BUFFER &&
> � � � �obj->buffer.length == sizeof(struct wmab_ret)) {
> � � � � � � � �ret = *((struct wmab_ret *) obj->buffer.pointer);
> � � � �} else {
> - � � � � � � � kfree(out.pointer);
> - � � � � � � � return AE_ERROR;
> + � � � � � � � status = AE_ERROR;
> + � � � � � � � goto out;
> � � � �}
>
> � � � �if (ret.eax & 0x1)
> @@ -614,25 +614,21 @@ static acpi_status AMW0_set_capabilities
> � � � � * But we need to kfree(out.pointer) if next wmab_execute fail.
> � � � � */
> � � � �status = wmab_execute(&args, &out);
> - � � � if (ACPI_FAILURE(status)) {
> - � � � � � � � kfree(out.pointer);
> - � � � � � � � return status;
> - � � � }
> + � � � if (ACPI_FAILURE(status))
> + � � � � � � � goto out;
>
> � � � �obj = (union acpi_object *) out.pointer;
> � � � �if (obj && obj->type == ACPI_TYPE_BUFFER
> � � � �&& obj->buffer.length == sizeof(struct wmab_ret)) {
> � � � � � � � �ret = *((struct wmab_ret *) obj->buffer.pointer);
> � � � �} else {
> - � � � � � � � kfree(out.pointer);
> - � � � � � � � return AE_ERROR;
> + � � � � � � � status = AE_ERROR;
> + � � � � � � � goto out;
> � � � �}
>
> � � � �if (ret.eax & 0x1)
> � � � � � � � �interface->capability |= ACER_CAP_BLUETOOTH;
>
> - � � � kfree(out.pointer);
> -
> � � � �/*
> � � � � * This appears to be safe to enable, since all Wistron based laptops
> � � � � * appear to use the same EC register for brightness, even if they
> @@ -641,7 +637,10 @@ static acpi_status AMW0_set_capabilities
> � � � �if (quirks->brightness >= 0)
> � � � � � � � �interface->capability |= ACER_CAP_BRIGHTNESS;
>
> - � � � return AE_OK;
> + � � � status = AE_OK;
> +out:
> + � � � kfree(out.pointer);
> + � � � return status;
> �}
>
> �static struct wmi_interface AMW0_interface = {
> _
>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: tracing: Add alignment to syscall metadata declarations
Next: [PATCH] arm: Build with -fno-dwarf2-cfi-asm