| commit | 50255feeaabcc7e7db22b858a6bd64a9b5b4f16d | |
| author | Lasse Collin <lasse.collin@tukaani.org> | |
| Mon, 22 Jan 2024 22:09:48 +0000 (23 00:09 +0200) | ||
| committer | Jia Tan <jiat0218@gmail.com> | |
| Tue, 23 Jan 2024 15:05:47 +0000 (23 23:05 +0800) | ||
| tree | bed06df927f6a9d1d91dd3360925ea3c1a3625bf | treesnapshot (tar.gz zip) |
| parent | db5eb5f563e8baa8d912ecf576f53391ff861596 | commitdiff |
liblzma: RISC-V filter: Use byte-by-byte access.
Not all RISC-V processors support fast unaligned access so
it's better to read only one byte in the main loop. This can
be faster even on x86-64 when compared to reading 32 bits at
a time as half the time the address is only 16-bit aligned.
The downside is larger code size on archs that do support
fast unaligned access.
Not all RISC-V processors support fast unaligned access so
it's better to read only one byte in the main loop. This can
be faster even on x86-64 when compared to reading 32 bits at
a time as half the time the address is only 16-bit aligned.
The downside is larger code size on archs that do support
fast unaligned access.
| src/liblzma/simple/riscv.c | diffblobblamehistory |