diff options
| author | Lennert Buytenhek <buytenh@wantstofly.org> | 2005-11-07 21:12:07 +0000 |
|---|---|---|
| committer | Russell King <rmk+kernel@arm.linux.org.uk> | 2005-11-07 21:12:07 +0000 |
| commit | 06c03cac9487555478c7d80065ebf7818bf6fd06 (patch) | |
| tree | 1a03ff9ae347f40c11d0e3eee790a1b88ff29cc1 /arch/arm/nwfpe/softfloat.h | |
| parent | 862184fe013146a0d9654a5598c5a2691747541c (diff) | |
[ARM] 3117/1: nwfpe kernel memory info leak
Patch from Lennert Buytenhek
The routine that nwfpe uses for converting floats/doubles to
extended precision fails to zero two bytes of kernel stack. This
is not immediately obvious, as the floatx80 structure has 16 bits
of implicit padding (by design.) These two bytes are copied to
userspace when an stfe is emulated, causing a possible info leak.
Make the padding explicit and zero it out in the relevant places.
Signed-off-by: Lennert Buytenhek <buytenh@wantstofly.org>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Diffstat (limited to 'arch/arm/nwfpe/softfloat.h')
| -rw-r--r-- | arch/arm/nwfpe/softfloat.h | 1 |
1 files changed, 1 insertions, 0 deletions
diff --git a/arch/arm/nwfpe/softfloat.h b/arch/arm/nwfpe/softfloat.h index 14151700b6b2..1301d97e037f 100644 --- a/arch/arm/nwfpe/softfloat.h +++ b/arch/arm/nwfpe/softfloat.h @@ -55,6 +55,7 @@ typedef unsigned long int float32; typedef unsigned long long float64; typedef struct { unsigned short high; + unsigned short __padding; unsigned long long low; } floatx80; |