liblzma: Omit zero-skipping from ARM64 filter.

It has some complicated downsides and its usefulness is more limited
than I originally thought. So this change is bad for certain very
specific situations but a generic solution that works for other
filters (and is otherwise better too) is planned anyway. And this
way 7-Zip can use the same compatible filter for the .7z format.

This is still marked as experimental with a new temporary Filter ID.
This commit is contained in:
Lasse Collin 2022-12-01 18:51:52 +02:00
parent 5baec3f0a9
commit f9ca7d4516
2 changed files with 24 additions and 59 deletions

View File

@ -49,7 +49,7 @@
* Filter for SPARC binaries. * Filter for SPARC binaries.
*/ */
#define LZMA_FILTER_ARM64 LZMA_VLI_C(0x3FDB87B33B27010B) #define LZMA_FILTER_ARM64 LZMA_VLI_C(0x3FDB87B33B27020B)
/**< /**<
* Filter for ARM64 binaries. * Filter for ARM64 binaries.
* *

View File

@ -6,13 +6,6 @@
/// This converts ARM64 relative addresses in the BL and ADRP immediates /// This converts ARM64 relative addresses in the BL and ADRP immediates
/// to absolute values to increase redundancy of ARM64 code. /// to absolute values to increase redundancy of ARM64 code.
/// ///
/// Unlike the older BCJ filters, this handles zeros specially. This way
/// the filter won't be counterproductive on Linux kernel modules, object
/// files, and static libraries where the immediates are all zeros (to be
/// filled later by a linker). Usually this has no downsides but with bad
/// luck it can reduce the effectiveness of the filter and trying a different
/// start offset can mitigate the problem.
///
/// Converting B or ADR instructions was also tested but it's not useful. /// Converting B or ADR instructions was also tested but it's not useful.
/// A majority of the jumps for the B instruction are very small (+/- 0xFF). /// A majority of the jumps for the B instruction are very small (+/- 0xFF).
/// These are typical for loops and if-statements. Encoding them to their /// These are typical for loops and if-statements. Encoding them to their
@ -30,20 +23,6 @@
#include "simple_private.h" #include "simple_private.h"
static uint32_t
arm64_conv(uint32_t src, uint32_t pc, uint32_t mask, bool is_encoder)
{
if (!is_encoder)
pc = 0U - pc;
uint32_t dest = src + pc;
if ((dest & mask) == 0)
dest = pc;
return dest;
}
static size_t static size_t
arm64_code(void *simple lzma_attribute((__unused__)), arm64_code(void *simple lzma_attribute((__unused__)),
uint32_t now_pos, bool is_encoder, uint32_t now_pos, bool is_encoder,
@ -51,29 +30,15 @@ arm64_code(void *simple lzma_attribute((__unused__)),
{ {
size_t i; size_t i;
// Clang 14.0.6 on x86-64 makes this four times bigger and 60 % slower // Clang 14.0.6 on x86-64 makes this four times bigger and 40 % slower
// with auto-vectorization that is enabled by default with -O2. // with auto-vectorization that is enabled by default with -O2.
// Even -Os, which doesn't use vectorization, produces faster code.
// Disabling vectorization with -O2 gives good speed (faster than -Os)
// and reasonable code size.
//
// Such vectorization bloat happens with -O2 when targeting ARM64 too // Such vectorization bloat happens with -O2 when targeting ARM64 too
// but performance hasn't been tested. // but performance hasn't been tested.
//
// Clang 14 and 15 won't auto-vectorize this loop if the condition
// for ADRP is replaced with the commented-out version. However,
// at least Clang 14.0.6 doesn't generate as fast code with that
// condition. The commented-out code is also bigger.
//
// GCC 12.2 on x86-64 with -O2 produces good code with both versions
// of the ADRP if-statement although the single-branch version is
// slightly faster and smaller than the commented-out version.
// Speed is similar to non-vectorized clang -O2.
#ifdef __clang__ #ifdef __clang__
# pragma clang loop vectorize(disable) # pragma clang loop vectorize(disable)
#endif #endif
for (i = 0; i + 4 <= size; i += 4) { for (i = 0; i + 4 <= size; i += 4) {
const uint32_t pc = (uint32_t)(now_pos + i); uint32_t pc = (uint32_t)(now_pos + i);
uint32_t instr = read32le(buffer + i); uint32_t instr = read32le(buffer + i);
if ((instr >> 26) == 0x25) { if ((instr >> 26) == 0x25) {
@ -87,27 +52,17 @@ arm64_code(void *simple lzma_attribute((__unused__)),
// so this is a compromise that slightly favors big // so this is a compromise that slightly favors big
// files. With the full range only six bits of the 32 // files. With the full range only six bits of the 32
// need to match to trigger a conversion. // need to match to trigger a conversion.
const uint32_t mask26 = 0x03FFFFFF; const uint32_t src = instr;
const uint32_t src = instr & mask26;
instr = 0x94000000; instr = 0x94000000;
if (src == 0) pc >>= 2;
continue; if (!is_encoder)
pc = 0U - pc;
instr |= arm64_conv(src, pc >> 2, mask26, is_encoder) instr |= (src + pc) & 0x03FFFFFF;
& mask26;
write32le(buffer + i, instr); write32le(buffer + i, instr);
/* } else if ((instr & 0x9F000000) == 0x90000000) {
// This is a more readable version of the one below but this
// has two branches. It results in bigger and slower code.
} else if ((instr & 0x9FF00000) == 0x90000000
|| (instr & 0x9FF00000) == 0x90F00000) {
*/
// This is only a rotation, addition, and testing that
// none of the bits covered by the bitmask are set.
} else if (((((instr << 8) | (instr >> 24))
+ (0x10000000 - 0x90)) & 0xE000009F) == 0) {
// ADRP instruction: // ADRP instruction:
// Only values in the range +/-512 MiB are converted. // Only values in the range +/-512 MiB are converted.
// //
@ -120,15 +75,25 @@ arm64_code(void *simple lzma_attribute((__unused__)),
// range, nine bits of 32 need to match to trigger a // range, nine bits of 32 need to match to trigger a
// conversion (two 10-bit match choices = 9 bits). // conversion (two 10-bit match choices = 9 bits).
const uint32_t src = ((instr >> 29) & 3) const uint32_t src = ((instr >> 29) & 3)
| ((instr >> 3) & 0x0003FFFC); | ((instr >> 3) & 0x001FFFFC);
instr &= 0x9000001F;
if (src == 0) // With the addition only one branch is needed to
// check the +/- range. This is usually false when
// processing ARM64 code so branch prediction will
// handle it well in terms of performance.
//
//if ((src & 0x001E0000) != 0
// && (src & 0x001E0000) != 0x001E0000)
if ((src + 0x00020000) & 0x001C0000)
continue; continue;
const uint32_t dest = arm64_conv( instr &= 0x9000001F;
src, pc >> 12, 0x3FFFF, is_encoder);
pc >>= 12;
if (!is_encoder)
pc = 0U - pc;
const uint32_t dest = src + pc;
instr |= (dest & 3) << 29; instr |= (dest & 3) << 29;
instr |= (dest & 0x0003FFFC) << 3; instr |= (dest & 0x0003FFFC) << 3;
instr |= (0U - (dest & 0x00020000)) & 0x00E00000; instr |= (0U - (dest & 0x00020000)) & 0x00E00000;