diff options
| author | Michael Brown <mcb30@ipxe.org> | 2018-03-20 20:42:39 +0200 |
|---|---|---|
| committer | Michael Brown <mcb30@ipxe.org> | 2018-03-20 20:56:01 +0200 |
| commit | 0d35411f88dd37dda755d52b4549337f8299c698 (patch) | |
| tree | 9129fb0cee236f97af83024b5910cb2cb2b83b08 /src/include/ipxe/linux | |
| parent | d5d4bf8870ccb68b3b897d2ee88079dc06e9dd48 (diff) | |
| download | ipxe-0d35411f88dd37dda755d52b4549337f8299c698.tar.gz | |
[rng] Use fixed-point calculations for min-entropy quantities
We currently perform various min-entropy calculations using build-time
floating-point arithmetic. No floating-point code ends up in the
final binary, since the results are eventually converted to integers
and asserted to be compile-time constants.
Though this mechanism is undoubtedly cute, it inhibits us from using
"-mno-sse" to prevent the use of SSE registers by the compiler.
Fix by using fixed-point arithmetic instead.
Signed-off-by: Michael Brown <mcb30@ipxe.org>
Diffstat (limited to 'src/include/ipxe/linux')
| -rw-r--r-- | src/include/ipxe/linux/linux_entropy.h | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/include/ipxe/linux/linux_entropy.h b/src/include/ipxe/linux/linux_entropy.h index afef6fe19..ea8c1f16c 100644 --- a/src/include/ipxe/linux/linux_entropy.h +++ b/src/include/ipxe/linux/linux_entropy.h @@ -20,7 +20,7 @@ FILE_LICENCE ( GPL2_OR_LATER_OR_UBDL ); * * @ret min_entropy min-entropy of each sample */ -static inline __always_inline double +static inline __always_inline min_entropy_t ENTROPY_INLINE ( linux, min_entropy_per_sample ) ( void ) { /* linux_get_noise() reads a single byte from /dev/random, @@ -28,7 +28,7 @@ ENTROPY_INLINE ( linux, min_entropy_per_sample ) ( void ) { * entropy is available. We therefore assume that each sample * contains exactly 8 bits of entropy. */ - return 8.0; + return MIN_ENTROPY ( 8.0 ); } #endif /* _IPXE_LINUX_ENTROPY_H */ |