diff options
| author | Daniel Jacobowitz <drow@false.org> | 2006-08-27 13:42:14 +0200 |
|---|---|---|
| committer | Russell King <rmk+kernel@arm.linux.org.uk> | 2006-08-27 13:42:14 +0200 |
| commit | b53a2b41f156a9c9b62c14502037cbc15bc08b54 (patch) | |
| tree | 7813d3dfbbb9fe97a8a093c837d7ee5a6746c4ae /arch/arm/vfp/vfpdouble.c | |
| parent | [ARM] 3749/3: Correct VFP single/double conversion emulation (diff) | |
| download | linux-b53a2b41f156a9c9b62c14502037cbc15bc08b54.tar.xz linux-b53a2b41f156a9c9b62c14502037cbc15bc08b54.zip | |
[ARM] 3758/1: Preserve signalling NaNs in conversion
Patch from Daniel Jacobowitz
The fcvtds and fcvtsd instructions were generating a qnan bit pattern
for both quiet and signalling NaNs.
Signed-off-by: Daniel Jacobowitz <dan@codesourcery.com>
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Diffstat (limited to 'arch/arm/vfp/vfpdouble.c')
| -rw-r--r-- | arch/arm/vfp/vfpdouble.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/arch/arm/vfp/vfpdouble.c b/arch/arm/vfp/vfpdouble.c index 021581680e94..04bd3425b29b 100644 --- a/arch/arm/vfp/vfpdouble.c +++ b/arch/arm/vfp/vfpdouble.c @@ -465,7 +465,7 @@ static u32 vfp_double_fcvts(int sd, int unused, int dm, u32 fpscr) */ if (tm & (VFP_INFINITY|VFP_NAN)) { vsd.exponent = 255; - if (tm & VFP_NAN) + if (tm == VFP_QNAN) vsd.significand |= VFP_SINGLE_SIGNIFICAND_QNAN; goto pack_nan; } else if (tm & VFP_ZERO) |