summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Revert "libc: Add NEON optimized memmove"Ricardo Cerqueira2013-11-0136-5637/+623
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This reverts commit e1154df8806b65a7706e1ba3e7b2a9f0fe2f32d0. Revert "libc: Optimize strcat/strcpy, small tweaks to strlen" This reverts commit dd2150b68a0e00d130bcf9b3b4af68a93971ae00. Revert "libc: Update to latest cortexa15 memcpy code." This reverts commit 392d2cd535185b4013ae7279587aaae3972737d1. Revert "Bionic: Add AOSP master cortex optimizations" This reverts commit 456c0e8cbe4ed2da25317e8e46ed327a14fbeb1c. Revert "libc: krait: Restore prior version of strcmp." This reverts commit d22231e03e36a34a149a90d2c01d867d62932f7f. Revert "libc: krait: Use performance version of memcpy." This reverts commit cc8a530aba0f1b616becbc0bbd40501f082d411f. Revert "libc: krait: Use performance version of bcopy and memmove" This reverts commit 9ef0a0f6002a3c0e0bcfa2b329811d9b3b8a534b. Revert "tegra2: add missing strcpy and strcat" This reverts commit 31a06f332f1176ae8e2a241dcae00f78112a3b2a. Revert "libc: add tegra2 TARGET_CPU_VARIANT" This reverts commit 4172806c7ee309fa78e20bb03313d0e8bc6bde0e. Revert "Detect userspace memory leak for JB_MR2" This reverts commit bba493121c8e135bd9cd145ab35e53cac1702a9e. Revert "[libc] Add missing strcat.S to x86.mk Makefile" This reverts commit b59b790f97dc58a931719524499269f2f3b904f2. Revert "armv6: REX routines are currently broken, use alternatives" This reverts commit 364c9574c789b046423f25903c9c2f36c1a32d01. Revert "Bump the number of TLS slots to 128." This reverts commit 5bce86fd740e13f4f6a55092d159af0200c46e72. Revert "don't hardcode register r0/v1 when reading the TLS" This reverts commit fb0f92fa912480e381c4a96ddd561e2360d6e1ca.
* [libc] Add missing strcat.S to x86.mk Makefilecm-10.2-M1Trevor Drake2013-09-271-0/+1
| | | | | | | arch-x86/string/strcat.S was not included in the x86.mk ; This was causing the build to fail when building for x86 Targets Change-Id: I8001cdb25c2ff84c994603e648ef74c9c1daa5ea
* add cortex-a5 cpu variantsakindia1232013-08-252-1/+2
| | | | Change-Id: I3f72fa4524c09d3ea9abcc3a4a693f563e58ed53
* Bump the number of TLS slots to 128.Elliott Hughes2013-08-251-1/+1
| | | | | Bug: 9997352 Change-Id: I7bde7228d803e9d4bb83309c5891d54a07e3b025
* Detect userspace memory leak for JB_MR2Maunik Shah2013-08-252-6/+173
| | | | | | | | | | | | | When process is consuming memory beyond the limit specified by property, it will print all the allocations of the same process. To enable this feature, do following steps: 1. libc.debug.malloc 40 2. libc.debug.malloc.program <PROCESS_NAME> 3. libc.debug.malloc.maxprocsize <VALUE_IN_BYTES> 4. libc.malloc.minalloclim <VALUE_IN_BYTES> Change-Id: I03a4de9643ec954802b26443ce5685975ea30f89 Signed-off-by: Maunik Shah <mshah@codeaurora.org>
* tegra2: add missing strcpy and strcatTanguy Pruvot2013-08-202-0/+136
| | | | | | | | | These applets are no more common since recent optimisations strcpy from cortex-a9 strcat : keep standard one Change-Id: Icfedf17aa4f9fa30e7f63b19bd6d4ac533e5e0e9
* don't hardcode register r0/v1 when reading the TLSMathias Agopian2013-08-201-12/+9
| | | | | | this leads to much improved code when calling __get_tls() Change-Id: I21d870fb33c33a921ca55c4e100772e0f7a8d1e4
* libc: add tegra2 TARGET_CPU_VARIANTTanguy Pruvot2013-08-147-1/+1145
| | | | | | | | | | | | | Tegra2 is a Dual Cortex A9 without FP neon SIMD based on 'generic' arm (memcpy, memset) on cortex-a15 and a8 (strcmp) and armv7-a from cm10.1 (memchr and strlen) rebased and now for cm-10.2 only Change-Id: I1e7207e2d1ba02012fb5306f46b688903d31386c Signed-off-by: Tanguy Pruvot <tanguy.pruvot@gmail.com>
* libc: Add NEON optimized memmoveathurh2013-08-133-3/+200
| | | | | | | | | | | | | | | | | | | | | | | | | | | ported from cm10.1 bionic-benchmarks on cortex-a9 generic: BM_string_memmove/8 50000000 67 119.35 MiB/s BM_string_memmove/64 10000000 175 364.59 MiB/s BM_string_memmove/512 1000000 1078 474.86 MiB/s BM_string_memmove/1K 1000000 2072 494.20 MiB/s BM_string_memmove/8K 100000 16400 499.50 MiB/s BM_string_memmove/16K 50000 32293 507.35 MiB/s BM_string_memmove/32K 50000 66585 492.12 MiB/s BM_string_memmove/64K 10000 160435 408.49 MiB/s NEON-optimized: BM_string_memmove/8 100000000 25 319.06 MiB/s BM_string_memmove/64 50000000 43 1472.60 MiB/s BM_string_memmove/512 10000000 247 2069.74 MiB/s BM_string_memmove/1K 5000000 463 2210.08 MiB/s BM_string_memmove/8K 500000 3465 2363.69 MiB/s BM_string_memmove/16K 500000 6894 2376.30 MiB/s BM_string_memmove/32K 100000 15490 2115.38 MiB/s BM_string_memmove/64K 50000 42097 1556.75 MiB/s Change-Id: I89253a01fb811438089e16320ac265177a2ca152
* Fix assembler errors in generic arm strlen.c.Christopher Ferris2013-08-131-4/+13
| | | | | | | | | | | Tested using a static version of the strlen libc_test program on a nexus7 that uses the generic code. Merge from internal master. (cherry-picked from d8d10a8994472e40d19301b7087806630877b4d5) Change-Id: I88f7dc01dc5b5c3ac2d5580d92153bc1bc36c564
* libc: Optimize strcat/strcpy, small tweaks to strlenChristopher Ferris2013-08-1314-38/+2250
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Create one version of strcat/strcpy/strlen for cortex-a15/krait/scorpion and another version for cortex-a9. Tested with the libc_test strcat/strcpy/strlen tests. Including new tests that verify that the src for strcat/strcpy do not overread across page boundaries. NOTE: The handling of unaligned strcpy (same code in strcat) could probably be optimized further such that the src is read 64 bits at a time instead of the partial reads occurring now. strlen improves slightly since it was recently optimized. Performance improvements for strcpy and strcat (using an empty dest string): cortex-a9 - Small copies vary from about 5% to 20% as the size gets above 10 bytes. - Copies >= 1024, about a 60% improvement. - Unaligned copies, from about 40% improvement. cortex-a15 - Most small copies exhibit a 100% improvement, a few copies only improve by 20%. - Copies >= 1024, about 150% improvement. - Unaligned copies, about 100% improvement. krait - Most small copies vary widely, but on average 20% improvement, then the performance gets better, hitting about a 100% improvement when copies 64 bytes of data. - Copies >= 1024, about 100% improvement. - When coping MBs of data, about 50% improvement. - Unaligned copies, about 90% improvement. As strcat destination strings get larger in size: cortex-a9 - about 40% improvement for small dst strings (>= 32). - about 250% improvement for dst strings >= 1024. cortex-a15 - about 200% improvement for small dst strings (>=32). - about 250% improvement for dst strings >= 1024. krait - about 25% improvement for small dst strings (>=32). - about 100% improvement for dst strings >=1024. Merge from internal master. (cherry-picked from d119b7b6f48fe507088cfb98bcafa99b320fd884) Change-Id: I296463b251ef9fab004ee4dded2793feca5b547a Conflicts: libc/Android.mk libc/arch-arm/generic/generic.mk libc/arch-arm/krait/krait.mk Fix strcpy.c that should have been strcpy.S Merge from internal master. (cherry-picked from 1ce665416307628f4bcaced86faa64bdf9c489c3) Conflicts: libc/arch-arm/generic/generic.mk Change-Id: I376b831df42248baadde7202a30a68112f752ff7
* libc: Update to latest cortexa15 memcpy code.Christopher Ferris2013-08-132-118/+354
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This uses the new code original submitted as memcpy.a15.S as the base. However, the old code handled unaligned src/dst better so that was spliced in. I optimized the original unaligned code by removing a few unnecessary instructions. I optimized the a15 code by rewriting the pre and post code. I also modified the main loop to add a pld so that larger copies would not stall waiting for memory. Test cases for the new memcpy: - Copy all sized values from 0 to 1024 bytes, using whatever alignment is returned by malloc. For each alignment case described below, the test copied from 0 to 128 bytes. - Src and dst pointers are both aligned to the same value, starting at one going through every power of two up to and including 128. - Src aligned to double word boundary, dst aligned to word boundary. - Src aligned to word boundary, dst aligned to double word boundary. - Src aligned to 16 bit boundary, dst aligned to word boundary. - Src aligned to word boundary, dst aligned to 16 byte boundary. - Src aligned to word boundary, dst aligned to 1 byte from a word boundary. - Src aligned to word boundary, dst aligned to 2 bytes from a word boundary. - Src aligned to word boundary, dst aligned to 3 bytes from a word boundary. - Src aligned to 1 byte from a word boundary, dst aligned to a word boundary. - Src aligned to 2 bytes from a word boundary, dst aligned to a word boundary. - Src aligned to 3 bytes from a word boundary, dst aligned to a word boundary. Cases to verify the unaligned source code properly aligns to a 16 bit boundary. - Src aligned to 1 byte from a 128 bit boundary, dst aligned to 4 + 128 bit boundary. - Src aligned to 1 byte from a 128 bit boundary, dst aligned to 8 + 128 bit boundary. - Src aligned to 1 byte from a 128 bit boundary, dst aligned to 12 + 128 bit boundary. - Src aligned to 1 byte from a 128 bit boundary, dst aligned to 16 + 128 bit boundary. In all cases, a two byte fencepost was placed at the end of the destination to verify that only the requested number of bytes were copied. Bug: 8005082 Merge from internal master. (cherry-picked from commit 21ede92d794969f22cacbdb9f557818f1c5712b5) Change-Id: Ief70c9e6dc8c6473ae245b6570b2c266fed9618c Add missing branch in memcpy.S dst aligned case. Merge from internal master. (cherry-picked from commit 6ffaa931c362602a2b606a610c92326a425a876e) Change-Id: Ifdcf01fd122866cf0d4c5b5f7a997803561d7889 Rewrite memset for cortexa15 to use strd. Merge from internal master. (cherry-picked from commit 7ffad9c120054eedebd5f56f8bed01144e93eafa) Change-Id: Ia67f2a545399f4fa37b63d5634a3565e4f5482f9
* libc: arch-arm: strlen: Illegal ARM entries fixedPaul Beeler2013-08-111-10/+10
| | | | | | | | | | | Spacing between brackets is a illegal ARM entry PS1 - Inital patch PS2 - update description and rebase Signed-off-by: Paul Beeler <sparksco@gmail.com> Change-Id: I52d7e95a8afb8b26f55fe0bc01853a15a95e90c2
* libc: Use the fast strlen on Scorpion tooSteve Kondik2013-08-101-0/+1
| | | | Change-Id: I9ba50a469de27269adc008c2a96767830ad6b70b
* Bionic: Add AOSP master cortex optimizationsWill Newton2013-08-1110-2/+805
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | libc/arch-arm/bionic/memcpy.a9.S: memcpy from cortex-strings. This memcpy code uses NEON/VFP to achieve very good performance on ARMv7-A processors. It is specifically tuned for A15 but should provide good performance on A9 also. It is equivalent to the code in cortex-strings rev 116. This patch is a follow up the existing gerrit change: I7f6f77995f3ca903ad9c66d14261441667a2a935 But this version includes a tweak for performance on misaligned buffers. Change-Id: I285abac0068f8ae29a1cbf7862ea8590aadaf0a7 Signed-off-by: Will Newton <will.newton@linaro.org> libc/arch-arm/bionic/memcpy.a9.S: memcpy from cortex-strings. This memcpy code uses NEON/VFP to achieve very good performance on ARMv7-A processors. It is specifically tuned for A15 but should provide good performance on A9 also. It is equivalent to the code in cortex-strings rev 116. This patch is a follow up the existing gerrit change: I7f6f77995f3ca903ad9c66d14261441667a2a935 This version includes a tweak for performance on misaligned buffers and splits the header comment into license and documentation sections. Change-Id: Ibd2e23c8d8e01357ba0247be1d05192de3ceba69 Signed-off-by: Will Newton <will.newton@linaro.org> Add new optimized strlen for arm. This optimized version is primarily targeted at cortex-a15. Tested on all nexus devices using the system/extras/libc_test strlen test. Tested alignments from 1 to 32 that are powers of 2. Tested that strlen does not cross page boundaries at all alignments. Speed improvements listed below: cortex-a15 - Sizes >= 32 bytes, ~75% improvement. - Sizes >= 1024 bytes, ~250% improvement. cortex-a9 - Sizes >= 32 bytes, ~75% improvement. - Sizes >= 1024 bytes, ~85% improvement. krait - Sizes >= 32 bytes, ~95% improvement. - Sizes >= 1024 bytes, ~160% improvement. Merge from internal master. (cherry-picked from 2fc071797743b88a9a47427d46baed7c7b24f4d2) Change-Id: I1ceceb4e745fd68e9d946f96d1d42e0cdaff6ccf Conflicts: libc/arch-arm/generic/generic.mk libc/arch-arm/krait/krait.mk libgcc_compat: Introduce __aeabi_lasr for cortex-a9 and higher This is needed when passing -mcpu=cortex-a9 or higher on a modern toolchain for prebuilt library compatibility Change-Id: I73eb2393377914ae26216a8c2828ad973d1c1225
* libc: Add Scorpion-optimized memcpy and memsetSteve Kondik2013-08-094-1/+229
| | | | | | | * Optimized routines from Qualcomm for msm8660 class SoCs * Enable with TARGET_CPU_VARIANT := scorpion Change-Id: I01d0f22efba5a418ddd20fca0d0c570d855e0f6f
* linker: restore prelink supportPawit Pornkitprasan2013-08-022-2/+40
| | | | | | | | | Prelink support is required to load old vendor binary blobs on many devices properly This commit partially reverts 4688279db5dcc4004941e7f133c4a1c3617d842c Change-Id: Iee5325e048daee78e8dcaa47c87454b908c89b4e
* libm: Fix multiple inclusion warningSteve Kondik2013-07-311-28/+22
| | | | Change-Id: Ifa732ec52fb582c34512cd1d09d918199ae7395e
* Bionic/libm: enable optimized math functions on more QCOM targets.Wei Wang2013-07-301-5/+15
| | | | | | allow more QCOM targets to use optimized math functions. Change-Id: I76ee1bf951ae1c8397fef3af6e9937ed8cad9b62
* Bionic/libm: Remove extra vmov from sin/cos.Lucas Crowthers2013-07-302-2/+4
| | | | | | | Move integer representations of x bits on the integer side rather than moving them to and from the FP registers. Change-Id: I1895db385c9616cdae9ab6403f392dfbae292adc
* libm: sqrt and sqrtf via ARM vsqrt instructionLucas Crowthers2013-07-303-5/+79
| | | | | | | | Optimized sqrt and sqrtf for arm by using hardware opcode for sqrt rather than generic slow portable code. Change-Id: I84694159577aef6418710548085d8149c45e0e3f
* libm: optimistic pow calling.Lucas Crowthers2013-07-303-18/+24
| | | | | | | Call optimized pow optimistically and revert to full range implementation if we detect an out-of-range input. Change-Id: I6f3aa734adbf99484b7ff70736ef83a41e5815b8
* Bionic/libm: Add precision-correct de-serialize sin/cos.Lucas Crowthers2013-07-302-0/+23
| | | | | | | Modify sin/cos to improve performance while retaining bit-for-bit accuracy with existing algorithm. Change-Id: Iaba2dd731cd015732744705dad8bddb713b43067
* Bionic/libm: Pow optimizations and bug fixes.Lucas Crowthers2013-07-302-33/+69
| | | | | | | | | Use VFP calling convention for pow_neon handoff function by default. Fix register usage collision between two different polynomial coefficients in pow_neon. Remove conditional execution in pow_neon and replace with branching. Change-Id: I76095f4a006e2fb01a53943b66fd69bfa1fd3033
* Bionic/libm: add assembly versions of sin/cos.Lucas Crowthers2013-07-303-3/+838
| | | | | | | | Add assembly versions of sin/cos with integrated remainder pi/2 calculation. Directly extracted from binary libm.so compiled with __ieee754_rem_pio2 calls inlined. Change-Id: Ia093f420e58e794635e3a5f09e8236ae7601f1f6
* Bionic/libm: Prefer branches and VFP ABI.Lucas Crowthers2013-07-302-0/+17
| | | | | | | | | For internal functions set gcc attribute "aapcs-vfp" for ARM and use -fno-if-conversion to prefer branches over predicated instructions (improves performance on architectures with good branch prediction). Change-Id: I8424e0e82a19d35e7e3b6e3e122dcdecdd5426fd
* Bionic/libm: fast neon pow() for small x,y.Lucas Crowthers2013-07-303-0/+444
| | | | | | | | | Add a fast neon version of pow() suitable for relatively small positive x and y (between 0 and 4). Run the standard implementation in all other cases. Gives approximately 60% performance improvement to AnTuTu FPU scores Change-Id: I97e0712daeb2740764b26a44be0caaa39c481453
* libc/arm: add cortex-a8 cpu variantRom Lemarchand2013-07-262-1/+2
| | | | Change-Id: I30e8dd6d4b2e7889aea8f5ed21182a5941bfb489
* libc: krait: Restore prior version of strcmp.Brent DeGraaf2013-07-251-365/+205
| | | | Change-Id: I8ddcdb4d3a905dd746985435dcdb525ab5a1c947
* libc: krait: Use performance version of bcopy and memmoveBrent DeGraaf2013-07-254-3/+215
| | | | Change-Id: I587085aeb0c30ceccaa3f420594a194b129632b5
* libc: krait: Use performance version of memcpy.Brent DeGraaf2013-07-251-108/+172
| | | | Change-Id: I47d25d1da5b1a96bbc1b60f8acdaa31721a68e73
* armv6: REX routines are currently broken, use alternativesPaul Mongold2013-07-241-4/+9
| | | | Change-Id: I7ac05ab6add7802a4cc24fe36f7181e7cdfe07e0
* headers: Unbreak some ARM syscallsRicardo Cerqueira2013-07-241-1/+1
| | | | | | | | Restore the outside-kernel exclusion for some syscalls that was removed by change I959b64280e184655ef8c713aa79f9e23cb1f7df4, since these syscalls are used elsewhere. Change-Id: I5b5bf3d78edd137e820d25281a375966b6c009ec
* getopt(): fix missing carriage return on bad parametersTanguy Pruvot2013-07-241-6/+6
| | | | | | | | This affect all command line tools which show the usage on stdout after an invalid parameter error (printed on stderr) Change-Id: I8e2cb3fda241ab989dc42055f15082f8b3ba1397 Signed-off-by: Tanguy Pruvot <tanguy.pruvot@gmail.com>
* Use GCC's __attribute__((const)) to reduce code sizeJim Huang2013-07-243-1/+3
| | | | | | | | | | | | | | | | | | | | | | __attribute__((const)) is mainly intended for the compiler to optimize away repeated calls to a function that the compiler knows will return the same value repeatedly. By adding __attribute__((const)), the compiler can choose to call the function just once and cache the return value. Therefore, this yields code size reduction. Here are the reference results by arm-eabi-size for crespo device: [before] text data bss dec hex filename 267715 10132 45948 323795 4f0d3 [after] text data bss dec hex filename 267387 10132 45948 323467 4ef8b Change-Id: I1d80465c0f88158449702d4dc6398a130eb77195
* merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-05-130-0/+0
|\
| * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-04-250-0/+0
| |\
| | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-04-110-0/+0
| | |\
| | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-04-080-0/+0
| | | |\
| | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-04-040-0/+0
| | | | |\
| | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-04-030-0/+0
| | | | | |\
| | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-04-020-0/+0
| | | | | | |\
| | | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-04-010-0/+0
| | | | | | | |\
| | | | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-03-280-0/+0
| | | | | | | | |\
| | | | | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-03-260-0/+0
| | | | | | | | | |\
| | | | | | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-03-250-0/+0
| | | | | | | | | | |\
| | | | | | | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-03-210-0/+0
| | | | | | | | | | | |\
| | | | | | | | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-03-180-0/+0
| | | | | | | | | | | | |\
| | | | | | | | | | | | | * am 2e7b8d63: am f861bc5c: Merge "Don\'t search off the end of the index for ↵Elliott Hughes2013-03-153-1/+52
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | bad Olson ids." * commit '2e7b8d6399fdea6e43dd07f353346324d2bf4ec4': Don't search off the end of the index for bad Olson ids.
| | | | | | | | | | | | | * merge in jb-mr2-release history after reset to jb-mr2-devThe Android Automerger2013-03-130-0/+0
| | | | | | | | | | | | | |\
generated by cgit v1.1 at 2025-02-10 23:26:28 +0000