Searched full:unrolled (Results 1 – 25 of 57) sorted by relevance
123
| /linux-6.14.4/tools/perf/bench/ |
| D | mem-memset-x86-64-asm-def.h | 4 "x86-64-unrolled",
5 "unrolled memset() in arch/x86/lib/memset_64.S")
|
| D | mem-memcpy-x86-64-asm-def.h | 4 "x86-64-unrolled",
5 "unrolled memcpy() in arch/x86/lib/memcpy_64.S")
|
| /linux-6.14.4/arch/alpha/lib/ |
| D | ev6-memset.S | 19 * however the loop has been unrolled to enable better memory throughput,
119 and $16, 0x3f, $2 # E : Forward work (only useful for unrolled loop)
126 * through unrolled loop. Do a quad at a time to get us 0mod64
194 * of data (where we can't use an unrolled loop and wh64)
297 and $16, 0x3f, $2 # E : Forward work (only useful for unrolled loop)
304 * through unrolled loop. Do a quad at a time to get us 0mod64
372 * of data (where we can't use an unrolled loop and wh64)
485 and $16, 0x3f, $2 # E : Forward work (only useful for unrolled loop)
492 * through unrolled loop. Do a quad at a time to get us 0mod64
560 * of data (where we can't use an unrolled loop and wh64)
|
| D | ev67-strchr.S | 66 * unrolled multiple times - is that safe to do in kernel space?
|
| D | memmove.S | 8 * memcpy whenever possible; the backwards copy loops are not unrolled.
|
| D | checksum.c | 78 * inner loop could be unrolled a bit further, and there are better
|
| /linux-6.14.4/drivers/gpu/drm/amd/amdgpu/ |
| D | gfx_v9_4_2_cleaner_shader.asm | 60 s_mov_b32 s2, 0x00000078 // Loop 128/8=16 times (loop unrolled for performance)
107 s_mov_b32 m0, 0x0000005c // Loop 96/4=24 times (loop unrolled for performance)
135 s_mov_b32 m0, 0x0000005c // Loop 96/4=24 times (loop unrolled for performance)
|
| D | gfx_v9_4_3_cleaner_shader.asm | 60 s_mov_b32 s2, 0x00000078 // Loop 128/8=16 times (loop unrolled for performance)
107 s_mov_b32 m0, 0x0000005c // Loop 96/4=24 times (loop unrolled for performance)
135 s_mov_b32 m0, 0x0000005c // Loop 96/4=24 times (loop unrolled for performance)
|
| D | gfx_v11_0_3_cleaner_shader.asm | 52 s_mov_b32 m0, 0x00000058 // Loop 96/8=12 times (loop unrolled for performance)
93 s_mov_b32 m0, 0x00000068 // Loop 108/4=27 times (loop unrolled for performance)
|
| D | gfx_v10_3_0_cleaner_shader.asm | 54 s_mov_b32 s2, 0x00000038 // Loop 64/8=8 times (loop unrolled for performance)
98 s_mov_b32 m0, 0x00000068 // Loop 108/4=27 times (loop unrolled for performance)
|
| /linux-6.14.4/arch/x86/lib/ |
| D | copy_user_uncached_64.S | 66 * just try the non-unrolled loop.
77 * falling back to the unrolled loop.
87 * depending on where in the unrolled loop
|
| D | clear_page_64.S | 120 * The unrolled case might end up clearing
|
| /linux-6.14.4/arch/riscv/lib/ |
| D | memmove.S | 45 * Both Copy Modes: a2 - Alternate breakpoint for unrolled loops
102 addi a2, t6, SZREG /* The other breakpoint for the unrolled loop*/
164 addi a2, t5, -SZREG /* The other breakpoint for the unrolled loop*/
|
| /linux-6.14.4/Documentation/devicetree/bindings/pci/ |
| D | snps,dw-pcie-ep.yaml | 34 if the space is unrolled (IP-core version >= 4.80a).
69 unrolled memory space with the internal Address Translation
|
| D | snps,dw-pcie.yaml | 43 also required if the space is unrolled (IP-core version >= 4.80a).
78 unrolled memory space with the internal Address Translation
|
| /linux-6.14.4/tools/perf/Documentation/ |
| D | perf-bench.txt | 189 On x86-64, x86-64-unrolled, x86-64-movsq and x86-64-movsb are supported.
213 On x86-64, x86-64-unrolled, x86-64-stosq and x86-64-stosb are supported.
|
| /linux-6.14.4/arch/m68k/lib/ |
| D | checksum.c | 57 /* unrolled loop for the main part: do 8 longs at once */ in csum_partial()
157 /* unrolled loop for the main part: do 8 longs at once */ in csum_and_copy_from_user()
292 /* unrolled loop for the main part: do 8 longs at once */ in csum_partial_copy_nocheck()
|
| /linux-6.14.4/lib/raid6/ |
| D | vpermxor.uc | 13 * $#-way unrolled portable integer math RAID-6 instruction set
|
| D | s390vx.uc | 5 * $#-way unrolled RAID6 gen/xor functions for s390
|
| D | altivec.uc | 16 * $#-way unrolled portable integer math RAID-6 instruction set
|
| D | int.uc | 16 * $#-way unrolled portable integer math RAID-6 instruction set
|
| D | mmx.c | 81 * Unrolled-by-2 MMX implementation
|
| /linux-6.14.4/arch/openrisc/lib/ |
| D | memcpy.c | 10 * unrolled implementation, but using word copies (if things are
|
| /linux-6.14.4/arch/x86/coco/ |
| D | core.c | 89 * With SEV, the rep string I/O instructions need to be unrolled in amd_cc_platform_has()
|
| /linux-6.14.4/Documentation/driver-api/mtd/ |
| D | nand_ecc.rst | 479 Unrolled the loop 1, 2, 3 and 4 times.
511 I decided to proceed with a four time unrolled loop anyway. It was my gut
539 compared with attempt 4 with 4 times unrolled, and we only require 1/3rd
674 Unrolled the loop one more time.
|
123