The ARM asm implementation of InternalMathSwapBytes64 () does interesting things if bit 7 of operand r1 (upper 32 bits of the input value) is set. After the recursive swap, bit 7 ends up in the sign bit position, after which it is right shifted with sign extension, and or'ed with the upper half of the output value.
This means SwapBytes64 (0x00000080_00000000) returns an incorrect value of 0xFFFFFFFF_80000000. Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <[email protected]> --- MdePkg/Library/BaseLib/Arm/Math64.S | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/MdePkg/Library/BaseLib/Arm/Math64.S b/MdePkg/Library/BaseLib/Arm/Math64.S index 4d975739207a..e2512621fe1a 100755 --- a/MdePkg/Library/BaseLib/Arm/Math64.S +++ b/MdePkg/Library/BaseLib/Arm/Math64.S @@ -256,22 +256,14 @@ L30: GCC_ASM_EXPORT(InternalMathSwapBytes64) ASM_PFX(InternalMathSwapBytes64): - @ args = 0, pretend = 0, frame = 0 - @ frame_needed = 1, uses_anonymous_args = 0 - stmfd sp!, {r4, r5, r6, r7, lr} - add r7, sp, #12 + stmfd sp!, {r4, r5, r7, lr} mov r5, r1 bl ASM_PFX(SwapBytes32) - mov r6, r0 + mov r4, r0 mov r0, r5 bl ASM_PFX(SwapBytes32) - mov r4, r6 - mov r5, r4, asr #31 - mov r2, #0 - mov r1, r0, asr #31 - orr r0, r0, r2 - orr r1, r1, r4 - ldmfd sp!, {r4, r5, r6, r7, pc} + mov r1, r4 + ldmfd sp!, {r4, r5, r7, pc} ASM_FUNCTION_REMOVE_IF_UNREFERENCED \ No newline at end of file -- 1.8.3.2 ------------------------------------------------------------------------------ Dive into the World of Parallel Programming The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net/ _______________________________________________ edk2-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/edk2-devel
