2022-11-16 15:32:09

by Shuah Khan

[permalink] [raw]
Subject: [PATCH] ACPICA: Fix sprintf() INT_MAX -Wformat-truncation= warn

utprint module is included in tools/acpidump and when acpidump is built,
the following warning occurs.

CC tools/acpidump/utprint.o
/usr/include/x86_64-linux-gnu/bits/stdio2.h: In function ‘sprintf’:
../../../../../drivers/acpi/acpica/utprint.c:602:18: warning: specified bound 4294901760 exceeds ‘INT_MAX’ [-Wformat-truncation=]
602 | length = vsnprintf(string, ACPI_UINT32_MAX-ACPI_UINT16_MAX,
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
603 | format, args);
| ~~~~~~~~~~~~~

Fix sprintf() to use ACPI_UINT32_MAX/2 to get rid of the truncate problem.
This change is inline other sprintf() implementations.

If using ACPI_UINT32_MAX is necessary in the kernel, acpidump might require
special case to get rid of this truncate problem.

Signed-off-by: Shuah Khan <[email protected]>
---
drivers/acpi/acpica/utprint.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/acpi/acpica/utprint.c b/drivers/acpi/acpica/utprint.c
index d5aa2109847f..02cff16c8f77 100644
--- a/drivers/acpi/acpica/utprint.c
+++ b/drivers/acpi/acpica/utprint.c
@@ -599,7 +599,7 @@ int sprintf(char *string, const char *format, ...)
int length;

va_start(args, format);
- length = vsnprintf(string, ACPI_UINT32_MAX, format, args);
+ length = vsnprintf(string, ACPI_UINT32_MAX/2, format, args);
va_end(args);

return (length);
--
2.34.1



2022-11-16 19:21:16

by Rafael J. Wysocki

[permalink] [raw]
Subject: Re: [PATCH] ACPICA: Fix sprintf() INT_MAX -Wformat-truncation= warn

On Wed, Nov 16, 2022 at 4:23 PM Shuah Khan <[email protected]> wrote:
>
> utprint module is included in tools/acpidump and when acpidump is built,
> the following warning occurs.
>
> CC tools/acpidump/utprint.o
> /usr/include/x86_64-linux-gnu/bits/stdio2.h: In function ‘sprintf’:
> ../../../../../drivers/acpi/acpica/utprint.c:602:18: warning: specified bound 4294901760 exceeds ‘INT_MAX’ [-Wformat-truncation=]
> 602 | length = vsnprintf(string, ACPI_UINT32_MAX-ACPI_UINT16_MAX,
> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> 603 | format, args);
> | ~~~~~~~~~~~~~
>
> Fix sprintf() to use ACPI_UINT32_MAX/2 to get rid of the truncate problem.
> This change is inline other sprintf() implementations.
>
> If using ACPI_UINT32_MAX is necessary in the kernel, acpidump might require
> special case to get rid of this truncate problem.

As ACPICA material, this should be submitted to the upstream project
on GitHub and honestly I'm not sure if it is going to be resolved this
way there.

> Signed-off-by: Shuah Khan <[email protected]>
> ---
> drivers/acpi/acpica/utprint.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/acpi/acpica/utprint.c b/drivers/acpi/acpica/utprint.c
> index d5aa2109847f..02cff16c8f77 100644
> --- a/drivers/acpi/acpica/utprint.c
> +++ b/drivers/acpi/acpica/utprint.c
> @@ -599,7 +599,7 @@ int sprintf(char *string, const char *format, ...)
> int length;
>
> va_start(args, format);
> - length = vsnprintf(string, ACPI_UINT32_MAX, format, args);
> + length = vsnprintf(string, ACPI_UINT32_MAX/2, format, args);
> va_end(args);
>
> return (length);
> --
> 2.34.1
>

2022-11-16 20:06:24

by Shuah Khan

[permalink] [raw]
Subject: Re: [PATCH] ACPICA: Fix sprintf() INT_MAX -Wformat-truncation= warn

On 11/16/22 12:05, Rafael J. Wysocki wrote:
> On Wed, Nov 16, 2022 at 4:23 PM Shuah Khan <[email protected]> wrote:
>>
>> utprint module is included in tools/acpidump and when acpidump is built,
>> the following warning occurs.
>>
>> CC tools/acpidump/utprint.o
>> /usr/include/x86_64-linux-gnu/bits/stdio2.h: In function ‘sprintf’:
>> ../../../../../drivers/acpi/acpica/utprint.c:602:18: warning: specified bound 4294901760 exceeds ‘INT_MAX’ [-Wformat-truncation=]
>> 602 | length = vsnprintf(string, ACPI_UINT32_MAX-ACPI_UINT16_MAX,
>> | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> 603 | format, args);
>> | ~~~~~~~~~~~~~
>>
>> Fix sprintf() to use ACPI_UINT32_MAX/2 to get rid of the truncate problem.
>> This change is inline other sprintf() implementations.
>>
>> If using ACPI_UINT32_MAX is necessary in the kernel, acpidump might require
>> special case to get rid of this truncate problem.
>
> As ACPICA material, this should be submitted to the upstream project
> on GitHub and honestly I'm not sure if it is going to be resolved this
> way there.

Thank you for your review. I will send this ACPICA upstream project.

thanks,
-- Shuah