cac5a7e871f1f346b317894359ad06fa7bd67fba |
|
22-Feb-2016 |
Vladimir Marko <vmarko@google.com> |
Optimizing: Improve const-string code generation. For strings in the boot image, use either direct pointers or pc-relative addresses. For other strings, use PC-relative access to the dex cache arrays for AOT and direct address of the string's dex cache slot for JIT. For aosp_flounder-userdebug: - 32-bit boot.oat: -692KiB (-0.9%) - 64-bit boot.oat: -948KiB (-1.1%) - 32-bit dalvik cache total: -900KiB (-0.9%) - 64-bit dalvik cache total: -3672KiB (-1.5%) (contains more files than the 32-bit dalvik cache) For aosp_flounder-userdebug forced to compile PIC: - 32-bit boot.oat: -380KiB (-0.5%) - 64-bit boot.oat: -928KiB (-1.0%) - 32-bit dalvik cache total: -468KiB (-0.4%) - 64-bit dalvik cache total: -1928KiB (-0.8%) (contains more files than the 32-bit dalvik cache) Bug: 26884697 Change-Id: Iec7266ce67e6fedc107be78fab2e742a8dab2696
|
208a5cb383dd9dcd3461f89b74af5df67dc8d794 |
|
03-Dec-2015 |
Mathieu Chartier <mathieuc@google.com> |
Store class tables in the image Reduces how long it takes to load an application image. N5 boot.art size Before: 8007680 After: 8122368 Also reduces boot time by how long AddImageClassesToClassTable used to take (~20ms). Changed class hashes to be uint32_t to fix cross compilation. We need serialized hash tables to be valid with different pointer sizes. Bug: 22858531 Change-Id: I463fc83f499ff75f509e80c253a55b9116ee5b89
|
1646d7a22e43a1fb25452ead47a4073e63d7f391 |
|
28-Oct-2015 |
Bruce Hoult <b.hoult@samsung.com> |
Optimize some commonly used utf8 functions by: - using counted loops instead of searching for terminating null. In the important cases the caller already knows the length: change the API to pass it in. Keep the old API version as well to avoid extensive changes to non-critical debug and test code. - ensure the common cases are at the start of if/then/else chains. Usually 99+% of characters are ASCII even in mixed strings. - for the "convert" functions, when both utf8 and utf16 lengths are passed, and are equal, it means the entire string is ASCII, and a specialized loop can be used. The compiler might then unroll or even vectorize this. The functions improved are (tested on Nexus 5 with a 44 character ASCII string): CountModifiedUtf8Chars : 20% faster ConvertUtf16ToModifiedUtf8: 80% faster ConvertModifiedUtf8ToUtf16: 200% faster Also for completeness CountUtf8Bytes has been cleaned up a little, but the speed is unchanged. Unlike CountModifiedUtf8Chars, it was already passed the length, rather than searching for null. Change-Id: I1c9b7dea3eda869fc9f5f6b4dd6be8cdd5bc3ac0
|
848f70a3d73833fc1bf3032a9ff6812e429661d9 |
|
15-Jan-2014 |
Jeff Hao <jeffhao@google.com> |
Replace String CharArray with internal uint16_t array. Summary of high level changes: - Adds compiler inliner support to identify string init methods - Adds compiler support (quick & optimizing) with new invoke code path that calls method off the thread pointer - Adds thread entrypoints for all string init methods - Adds map to verifier to log when receiver of string init has been copied to other registers. used by compiler and interpreter Change-Id: I797b992a8feb566f9ad73060011ab6f51eb7ce01
|
e16dad1d6388b0305f13e2171308a77f42e7c682 |
|
13-Feb-2015 |
Narayan Kamath <narayan@google.com> |
Emit 4 byte UTF-sequences in place of encoded surrogate pairs. Symmetric with a5afcfc73141e5e378d79a326d0 which converts 4 byte UTF-8 sequences to surrogate pairs. bug: 18848397 Change-Id: I42adc275b7e0df0cbbd9d8a799e8b0447d8f5cae
|
a5afcfc73141e5e378d79a326d02c5c2039fb025 |
|
29-Jan-2015 |
Narayan Kamath <narayan@google.com> |
Be more lenient with 4 byte UTF-8 sequences. Accept 4 byte sequences and convert them into surrogate pairs instead of expecting 2 separate 3 byte sequences each encoding one half of a surrogate pair. Note that in addition to supporting 4 byte sequences in strings from JNI, we also tolerate them in dex files. This is mainly for consistency, and there's no need to claim any sort of official support. bug: 18848397 bug: https://code.google.com/p/android/issues/detail?id=81341 Change-Id: Ibc98d29e59d98803e640f2489ea4c56912a59b29
|
a48aef4234768ed37828df613919391c21f561a7 |
|
03-Dec-2014 |
Vladimir Marko <vmarko@google.com> |
Fix ImageWriter::ComputeEagerResolvedStringsCallback(). Change-Id: I1a2abd6d78dd7067d9bdbadbd81dd2fd7711fbc5
|
e7c9a8c2b8481aafbc6af4ce6229bd361ba24742 |
|
07-Nov-2014 |
Mathieu Chartier <mathieuc@google.com> |
Add hash map, reduce excessive hashing Changed the class def index to use a HashMap instead of unordered_map so that we can use FindWithHash to reduce how often we need to compute hashes. Fixed a bug in ClassLinker::UpdateClass where we didn't properly handle classes with the same descriptor but different class loaders. Introduced by previous CL. Before (fb launch): 1.74% art::ComputeModifiedUtf8Hash(char const*) After: 0.95% art::ComputeModifiedUtf8Hash(char const*) Bug: 18054905 Bug: 16828525 Change-Id: Iba2ee37c9837289e0ea187800ba4af322225a994 (cherry picked from commit 564ff985184737977aa26c485d0c1a413e530705)
|
564ff985184737977aa26c485d0c1a413e530705 |
|
07-Nov-2014 |
Mathieu Chartier <mathieuc@google.com> |
Add hash map, reduce excessive hashing Changed the class def index to use a HashMap instead of unordered_map so that we can use FindWithHash to reduce how often we need to compute hashes. Fixed a bug in ClassLinker::UpdateClass where we didn't properly handle classes with the same descriptor but different class loaders. Introduced by previous CL. Before (fb launch): 1.74% art::ComputeModifiedUtf8Hash(char const*) After: 0.95% art::ComputeModifiedUtf8Hash(char const*) Bug: 18054905 Bug: 16828525 Change-Id: Iba2ee37c9837289e0ea187800ba4af322225a994
|
8f41dc389bd50b6e31caa3fe9d41cbaf76e4d8d6 |
|
30-Oct-2014 |
Ian Rogers <irogers@google.com> |
Avoid signed arithmetic overflow in UTF hash functions. Change-Id: I3485ef8445a6034661c1e75dc11ab9ad48884458
|
cc2f2393e69a9b1425bad1a89f41aaaf8c38f9e2 |
|
30-Aug-2014 |
Ian Rogers <irogers@google.com> |
Reduce and speed-up class def searches. Use the class linker for descriptor lookups from the compile driver so that dex caches are populated. Reduce the scope of functions for scanning class paths to just the class linker where they are performed. If we see more than a threshold number of find class def misses on a dex file lazily compute an index, so that future lookups are constant time (part of the collection code is taken from https://android-review.googlesource.com/#/c/103865/3). Note that we take a lazy approach so that we don't serialize on loading dex files, this avoids the reason the index was removed in 8b2c0b9abc3f520495f4387ea040132ba85cae69. Remove an implicit and unnecessary std::string creation for PrintableString. Single threaded interpret-only dex2oat performance is improved by roughly 10%. Bug: 16853450 Change-Id: Icf72df76b0a4328f2a24075e81f4ff267b9401f4 (cherry picked from commit 68b56858367e29461ae290fd797443a1ef6d8005)
|
68b56858367e29461ae290fd797443a1ef6d8005 |
|
30-Aug-2014 |
Ian Rogers <irogers@google.com> |
Reduce and speed-up class def searches. Use the class linker for descriptor lookups from the compile driver so that dex caches are populated. Reduce the scope of functions for scanning class paths to just the class linker where they are performed. If we see more than a threshold number of find class def misses on a dex file lazily compute an index, so that future lookups are constant time (part of the collection code is taken from https://android-review.googlesource.com/#/c/103865/3). Note that we take a lazy approach so that we don't serialize on loading dex files, this avoids the reason the index was removed in 8b2c0b9abc3f520495f4387ea040132ba85cae69. Remove an implicit and unnecessary std::string creation for PrintableString. Single threaded interpret-only dex2oat performance is improved by roughly 10%. Bug: 16853450 Change-Id: Icf72df76b0a4328f2a24075e81f4ff267b9401f4
|
ef7d42fca18c16fbaf103822ad16f23246e2905d |
|
06-Jan-2014 |
Ian Rogers <irogers@google.com> |
Object model changes to support 64bit. Modify mirror objects so that references between them use an ObjectReference value type rather than an Object* so that functionality to compress larger references can be captured in the ObjectRefererence implementation. ObjectReferences are 32bit and all other aspects of object layout remain as they are currently. Expand fields in objects holding pointers so they can hold 64bit pointers. Its expected the size of these will come down by improving where we hold compiler meta-data. Stub out x86_64 architecture specific runtime implementation. Modify OutputStream so that reads and writes are of unsigned quantities. Make the use of portable or quick code more explicit. Templatize AtomicInteger to support more than just int32_t as a type. Add missing, and fix issues relating to, missing annotalysis information on the mutator lock. Refactor and share implementations for array copy between System and uses elsewhere in the runtime. Fix numerous 64bit build issues. Change-Id: I1a5694c251a42c9eff71084dfdd4b51fff716822
|
a67249065e4c9b3cf4a7c081d95a78df28291ee9 |
|
23-Sep-2013 |
Ian Rogers <irogers@google.com> |
Move hot utf routines into -inl.h. Change-Id: I7050d8282a7e5870b2bf671d6867c57625e00ccc
|
7940e44f4517de5e2634a7e07d58d0fb26160513 |
|
12-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Create separate Android.mk for main build targets The runtime, compiler, dex2oat, and oatdump now are in seperate trees to prevent dependency creep. They can now be individually built without rebuilding the rest of the art projects. dalvikvm and jdwpspy were already this way. Builds in the art directory should behave as before, building everything including tests. Change-Id: Ic6b1151e5ed0f823c3dd301afd2b13eb2d8feb81
|