| 316a2186b7fa9e03187d45ac0fa320f4dff1f3df |
|
18-Sep-2015 |
Douglas Leung <douglas.leung@imgtec.com> |
Add Mips abiflags section to the oatfile.
Once this patch is merged, then the workaround patch:
0acb15ead6a554a6879b29fd90726b9ea8fd98c4
in bionic can be removed.
Change-Id: Ie06a3c4e384a23a77db7d04a2508edbf3a6d3933
|
| 5d8112029d0e085c5a0099257daa4c7e29c12310 |
|
08-Mar-2016 |
David Srbecky <dsrbecky@google.com> |
Propagate InstructionSetFeatures to ElfBuilder.
This is subset of CL171665 and it separates it into two.
It will be needed to generate .MIPS.abiflags ELF section.
Change-Id: I5557e7cb98d0fa1dc57c85cf6161e119c6d50a1a
|
| 944da603cde59a4277f3bbc31d860a90842a1a2a |
|
19-Feb-2016 |
Vladimir Marko <vmarko@google.com> |
ART: Allow method references across oat files for multi-image, 2nd attempt.
These were disabled because we didn't have sufficient
information about the multi-image layout when processing
link-time patches in OatWriter. This CL refactors the
ELF file creation so that the information is available.
Also clean up ImageWriter to use oat file indexes instead
of filenames and avoid reopening the oat file to retrieve
the checksum.
Change-Id: Icc7b528deca29da1e473c8f079521a36d6c4892f
|
| 45724f9a0cc38dbb3071beb3eeab96499868b49c |
|
17-Feb-2016 |
Vladimir Marko <vmarko@google.com> |
Revert "Allow method references across oat files for multi-image."
Breaks Quick tests.
This reverts commit 6065402316da2b51eed5fc34cffbd991766bd408.
Change-Id: I8a5469ba7cea5f46b85cb489b3e0ef06ed548f03
|
| 6065402316da2b51eed5fc34cffbd991766bd408 |
|
16-Feb-2016 |
Vladimir Marko <vmarko@google.com> |
Allow method references across oat files for multi-image.
These were disabled because we didn't have sufficient
information about the multi-image layout when processing
link-time patches in OatWriter. This CL refactors the
ELF file creation so that the information is available.
Change-Id: I6f2e8dc8572d143c72cc2693bb0ba4fd76f889ab
|
| c5bfa97c47d656b76f297af8abcd5f7502987399 |
|
05-Feb-2016 |
David Srbecky <dsrbecky@google.com> |
Split elf_writer_debug.cc to several files.
Refactoring only. The file has grown significantly over time,
and it is time to split it so it can be better managed.
Change-Id: Idce0231718add722292f4701df353d5baf31de5f
|
| 4fda4eb799c95be266f52aaf3461a440ea86b841 |
|
05-Feb-2016 |
David Srbecky <dsrbecky@google.com> |
Move code related to debug info generation to its own directory.
debug/dwarf/ contains helper classes which hide the details
of the DWARF file format. It acts as independent DWARF library.
debug/ contains ART-specific code which generates ELF debug
sections (which includes non-DWARF sections like .symtab).
Change-Id: Id351f604e4e64be2ca395a78324ea02e30481497
|
| 370339cb8edd91ec8732e0c97cea1f0556cfcc11 |
|
05-Feb-2016 |
David Srbecky <dsrbecky@google.com> |
Do not emit DWARF debug info if there are no methods.
Change-Id: I43406e54b454dbac45980d5c5edd90cd5593b9fd
|
| 0c4572e8c874de279463af22268d588471f40d3e |
|
22-Jan-2016 |
David Srbecky <dsrbecky@google.com> |
Generate mini-debug-info on separate thread.
The generation and compression of mini-debug-info is a lot of work.
However, we can do it on background thread when the main thread is
busy with .rodata and .text related I/O.
Change-Id: I514f1db3cb50aa250639f3ef697faa9bc9976d12
|
| ae5d2738a2b941b543c3fd478af910d4cd16f2ba |
|
05-Feb-2016 |
David Srbecky <dsrbecky@google.com> |
Change the method which generates DWARF mini-debug-info.
This splits some code from CL198651. It moves the WriteSection
call out one level and does not otherwise change behaviour.
Change-Id: I7dc1c7c08b577b50bf6fa366a9b0ca757048b81e
|
| 579942fc26f8a0b0ff1d909d19086fca6d2396a3 |
|
28-Jan-2016 |
David Srbecky <dsrbecky@google.com> |
Reduce the size of native debug info generated by JIT.
Remove some ELF file overheads:
- Do not produce program headers.
- Do not page align strtab.
- Do not write oat_patches sections.
This more than halves the size of JIT native debug info.
Since we generate many small entries, the overheads added up.
Change-Id: I27d95548c61e2e38c3683d6f5eb870a2db6e812d
|
| 5b1c2ca30dad519be285f0a1e839c23cc4e3a51d |
|
25-Jan-2016 |
David Srbecky <dsrbecky@google.com> |
Revert "Revert "Add option to generate compressed backtrace info.""
This reverts commit 8546cc9aeb05e866e1fb6a9e4130d53ea330baa8.
Change-Id: I676fdf9af27fa3b16fa8921778ff8832ab8c437d
|
| 8546cc9aeb05e866e1fb6a9e4130d53ea330baa8 |
|
25-Jan-2016 |
David Srbecky <dsrbecky@google.com> |
Revert "Add option to generate compressed backtrace info."
This reverts commit 5fdcc3c931b70204fd8c491afa66f57f8428490f.
Change-Id: I9c1f5aad6933a46af6717e3a90a51f76111f9c8a
|
| 5fdcc3c931b70204fd8c491afa66f57f8428490f |
|
19-Jan-2016 |
David Srbecky <dsrbecky@google.com> |
Add option to generate compressed backtrace info.
Add flag --generate-mini-debug-info which generates
LZMA compressed .symtab and .debug_frame, which are
sufficient to print java backtraces in libunwind.
If enabled, it increases the size of boot.oat by about 3.5%.
Change-Id: Ic3c2ef7704c05fa328720c6781ca2a9b8e3935a3
|
| fffbee4d158259633ec7b7f712eaf75be86bd4e5 |
|
15-Jan-2016 |
Tamas Berghammer <tberghammer@google.com> |
Report types loaded during init to the native debugger
The runtime loads a lot of type before it creates the jit from the
boot image and from hard coded source code (e.g. primitive arrays).
This change emits type information for these types after the jit has
been created. At the same time we remove the type info generation
that happens during AOT compilation because that type information can
be modified by a class loader at runtime.
Change-Id: Ie5b3b3df9d01c7200a1f670a98d9cbee796234e9
|
| 86e4278d5edc3685465b8846dcb17efa83c86d75 |
|
05-Jan-2016 |
Tamas Berghammer <tberghammer@google.com> |
Add DWARF type information generation.
Emit native debugging information for types which are used during
compilation.
Change-Id: If28d19f60294494b7c6db8400d179494bebe9e61
|
| e0febdfc2d9585d0d1b4a562b8e44d9058c71c37 |
|
17-Dec-2015 |
David Srbecky <dsrbecky@google.com> |
Move debug symbol writing code to elf_writer_debug.cc
I want to be able to use this code without needing to
reference elf_writer_quick.
Change-Id: I87ccedc540b8f04b53be0fc6344a5947308ffcb6
|
| 131980fc9aeb2b4d03480443e0fb494c76ce03a2 |
|
03-Dec-2015 |
Vladimir Marko <vmarko@google.com> |
Add ElfWriter::GetStream().
This will be used for writing the OatHeader which is
currently oddly written through the .text stream.
Also move the error delaying output stream out of the
ElfBuilder<> to its own file and move all output stream
files to compiler/linker/.
Change-Id: I00db4e33ed80ac4757ec459946c7b5ae014a3a2e
|
| 10c13565474de2786aad7c2e79757ea250747a15 |
|
25-Nov-2015 |
Vladimir Marko <vmarko@google.com> |
Refactor oat file writing to give Dex2Oat more control.
This is the first step towards writing dex files to oat file
and mapping them from there for the actual AOT compilation.
Change-Id: Icb0d27487eaf6ba3a66c157e695f9bdc5bb9cf9a
|
| 6d8c8f0344a706df651567387ede683ab3ec1b5f |
|
26-Oct-2015 |
David Srbecky <dsrbecky@google.com> |
Rewrite ElfBuilder to make streaming directly to file easier.
The previous design required knowing all the sections and their
sizes before even the first byte of the file was written.
The new design allows sections to be written one by one without
any knowledge of later sections. Furthermore, as soon as
section is started, its virtual memory address is known,
which removes the need for the various patching passes.
The new ElfBuilder essentially tries to be a thin wrapper
around OutputStream which keeps track where the various
sections start/end and then writes their ELF headers.
Change-Id: I817e7f3b41882e4e4b9b442cfe56e4ef2e26babd
|
| a26cb57f46fd3f27a930d9d688fe8670c1f24754 |
|
23-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
ART stack unwinding fixes for libunwind/gdb/lldb.
dex2oat can already generate unwinding and symbol information which
allows tools to create backtrace of mixed native and Java code.
This is a cherry pick from aosp/master which fixes several issues.
Most notably:
* It enables generation of ELF-64 on 64-bit systems (in dex2oat, C
compilers already produce ELF-64). Libunwind requires ELF-64 on
64-bit systems for backtraces to work.
* It enables loading of ELF files with dlopen. This is required for
libunwind to be able to generate backtrace of current process (i.e.
the process requesting backtrace of itself).
* It adds unit test to test the above (32 vs 64 bit, in-proces vs
out-of-process, application code vs framework code).
* Some other fixes or clean-ups which should not be of much
significance but which are easier to include to make the
important CLs cherry-pick cleanly.
This is squash of the following commits from aosp/master:
7381010 ART: CFI Test
e1bbed2 ART: Blacklist CFI test for non-compiled run-tests
aab9f73 ART: Blacklist CFI test for JIT
4437219 ART: Blacklist CFI test for Heap Poisoning
a3a49fe Switch to using ELF-64 for 64-bit architectures.
297ed22 Write 64-bit address in DWARF if we are on 64-bit architecture.
24981a1 Set correct size of PT_PHDR ELF segment.
1a146bf Link .dynamic to .dynstr
67a0653 Make some parts of ELF more (pointer) aligned.
f50fa82 Enable 64-bit CFI tests.
49e1fab Use dlopen to load oat files.
5dedb80 Add more logging output for dlopen.
aa03870 Find the dlopened file using address rather than file path.
82e73dc Release dummy MemMaps corresponding to dlopen.
5c40961 Test that we can unwind framework code.
020c543 Add more log output to the CFI test.
88da3b0 ART: Fix CFI test wrt/ PIC
a70e5b9 CFI test: kill the other process in native code.
ad5fa8c Support generation of CFI in .debug_frame format.
90688ae Fix build - large frame size of ElfWriterQuick<ElfTypes>::Write.
97dabb7 Fix build breakage in dwarf_test.
388d286 Generate just single ARM mapping symbol.
f898087 Split .oat_patches to multiple sections.
491a7fe Fix build - large frame size of ElfWriterQuick<ElfTypes>::Write (again).
8363c77 Add --generate-debug-info flag and remove the other two flags.
461d72a Generate debug info for core.oat files.
Bug: 21924613
Change-Id: I3f944a08dd2ed1df4d8a807da4fee423fdd35eb7
|
| 8363c772581bf00ebcdc2e38391b4bfae51beb75 |
|
28-May-2015 |
David Srbecky <dsrbecky@google.com> |
Add --generate-debug-info flag and remove the other two flags.
Replace the flags --include-debug-symbols and --include-cfi
with single flag called --generate-debug-info (with alias -g).
The name "symbol" was not ideal, since depending on context it
may be interpreted as "ELF symbols", or "debugging information".
This CL also means that we have only the options to include
either all debugging information or none. This should be fine,
since we can use standard tools to strip anything we do not want.
Change-Id: I721fded56d755d7eed0ef36aa84e841a1f5747f8
|
| 491a7fed21eb153965cee73ac77268ec340aaca2 |
|
28-May-2015 |
David Srbecky <dsrbecky@google.com> |
Fix build - large frame size of ElfWriterQuick<ElfTypes>::Write (again).
Change-Id: I8a3277835f22a23c31063eda9b7f1fdc9423a5b0
|
| f8980875ef8fb0ce86be4ed2c0af7070f5ae9cfd |
|
22-May-2015 |
David Srbecky <dsrbecky@google.com> |
Split .oat_patches to multiple sections.
.oat_patches section currently contains encoded patch locations for
several other sections. Split it to several sections - one for each
of the destination sections. For example, .text.oat_patches section
contains patch locations for the .text section.
This ensures that if we strip some the sections using standard
tools, we strip the corresponding .oat_patches section as well.
It also makes the overall design simpler.
I should have done it this way in the first place.
Since ApplyOatPatches has been simplified and uses unaligned memory
access, this also fixes bug 21403573.
Bug:20556771
Bug:21403573
Change-Id: Iae7c423911507b524eec500e8d61744046fcd3ba
|
| 388d2861ce185fe9bbf1989f1467031467bd1de7 |
|
21-May-2015 |
David Srbecky <dsrbecky@google.com> |
Generate just single ARM mapping symbol.
It is unnecessary to keep repeating the $t symbol if there
are no $d symbols. The last $t should still be in effect.
This shrinks the .symtab section by half.
Change-Id: Ic57c8c2d412c10f0d040e966379ec524ece87d4a
|
| aaf143d80bc6f226290eadb10361a752ed5d204e |
|
21-May-2015 |
David Srbecky <dsrbecky@google.com> |
Switch to using .debug_frame for CFI.
This will make it easier to strip or compress CFI using standard tools.
It also saves some space since we do not need .eh_frame_hdr anymore.
Bug:20556771
Change-Id: I8656b1ff248e4b12cb94924800c3a91fac8bdda4
|
| ad5fa8c5b26a325dc2a9521b87188755046c17f3 |
|
06-May-2015 |
David Srbecky <dsrbecky@google.com> |
Support generation of CFI in .debug_frame format.
.debug_frame section is almost identical to .eh_frame section.
There are only minor differences in the CIE and FDE headers.
The main difference is that .eh_frame is intended to be used
at runtime for exception handling and is therefore allocated
within the running program whereas .debug_frame is not.
This makes .debug_frame easier to remove using standard tools.
"objcopy --strip-debug" removes .debug_frame, but not .eh_frame.
(although objcopy can be still be forced to remove .eh_frame)
Similarly, we might want to separate the CFI to separate debug
file or include it as compressed .gnu_debugdata section.
It is more appropriate to use .debug_frame for this purpose.
Bug:20556771
Change-Id: I9d91a333b9fb37523fd6fafccfad89b21d2477af
|
| 033d745bb9412c9b546d29395cc9efbbb4fa306e |
|
30-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Add binary search table to .eh_frame_hdr
The addresses in the search table must be relative to the start
of .eh_frame_hdr which results in yet another pointer type to patch.
I have moved the Patch method back to elf_writer_quick where
it can be specialized to one of the three pointer types.
The .eh_frame_hdr takes around 17% of .eh_frame now.
Change-Id: I4770d1355ded6cdff9489c33380f6b06c4e3f9fe
|
| 6d73c9d06bc0fc6b32825ca0a8224010933a026e |
|
01-May-2015 |
David Srbecky <dsrbecky@google.com> |
Do not write CFI and symbols for deduplicate methods.
There is no need to have multiple symbols for a given address
since libunwind still has to pick only one to print.
Likewise, there is no need to duplicate the CFI.
There is fair number of duplicate methods in the boot image,
so this saves over 20% from those debug sections.
Change-Id: Ib4390150257d009a6303b084076750ce56afed60
|
| b0a962c59699fb4d115fb159eeabbd2200c6f872 |
|
28-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Move .eh_frame to the end of ELF file.
Move .eh_frame to the end together with other debug sections.
This should make it easier to remove it using standard strip tools.
Bug:20556771
Change-Id: I8303ccb053aee50d53fb193819d345c91f1e5b75
|
| bc90fd09e09a845ae6ea0d84ad67560575a94142 |
|
22-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Major refactoring of ElfBuilder.
Simplify ElfBuilder by removing duplicated or redundant code.
Many of the repeated code patterns were replaced by just
looping over the list of all sections.
Methods Init() and Write() have been merged into one.
The split between those was rather arbitrary, but it was there
for a reason. It allowed creation of raw sections between
the calls which may have depended on layout decisions done in
Init(), but not in Write() (e.g. knowing of offset of .text).
This has been replaced by more generic solution.
All sections are asked about their size first and complete
file layout is calculated. Then the sections are asked to
write their content (potentially using the layout information).
This should be pure refactoring CL - the compiler should
produce bit for bit identical output as before.
Change-Id: I281d13d469801bd8288b36b360d200d98a3e92d7
|
| 0c5bbc11a6af7b4f83b3929bc99de690635095c0 |
|
28-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Minor refactoring of ElfBuilder.
* Nest section builder classes within the main builder
class to reduce verbosity caused by templates.
* Move .eh_frame and .eh_frame_hdr section headers earlier
so that the section headers are sorted by file offset.
* Do not emit PT_NULL header if .eh_frame_hdr is missing.
None of these changes should have any material impact.
This CL just prepares ground for the following CL.
Change-Id: I56f9825892adb01d413a18ca10cb5c805cb8cd78
|
| 2cebb24bfc3247d3e9be138a3350106737455918 |
|
22-Apr-2015 |
Mathieu Chartier <mathieuc@google.com> |
Replace NULL with nullptr
Also fixed some lines that were too long, and a few other minor
details.
Change-Id: I6efba5fb6e03eb5d0a300fddb2a75bf8e2f175cb
|
| 533c207f9d2da6d913c4b10f6f757fe9d6367b10 |
|
22-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Simplify template parameters of Elf classes.
The ELF specification defines several types which differ between
32-bit ELF and 64-bit ELF. We used to template all ELF-related
methods on all of those types which was very verbose.
This CL wraps all the types as typedefs in ElfTypes32 and ElfTypes64.
One of those wrappers is then used as the template parameter.
Change-Id: I65247c2c79d92a7c4799e988cf3e4a1b10eb4788
|
| 527c9c71f0c6e2f9943ac028e7c050500699a44b |
|
17-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Generate .eh_frame_hdr section and PT_GNU_EH_FRAME segment.
Fixes issue 20125400 - ART: Need .eh_frame_hdr and PT_GNU_EH_FRAME
for libunwind.
.eh_frame_hdr serves two purposes. Firstly, it can optionally contain
binary search table for fast eh_frame lookup. This is important for
C++ exception handling, but we do not need it so we omit it.
Secondly, it contains a relative .eh_frame pointer which makes it
easier for run-time code to locate the .eh_frame section.
libunwind seems to rely on this relative pointer.
Bug: 20125400
Change-Id: I7c1e3f68d914f70781404c508395831a3296a7da
|
| 626a1666015b0fa201e979870baf06afa93b65e7 |
|
12-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Append [DEDUPED] suffix to all deduped methods in the symbol table.
This resolves old TODO in the code. The first method was never
marked as deduped and only the subsequent copies were recognised.
Therefore the suffix might have been missing in backtraces.
Change-Id: I4882d90f3049f7e196cd38c8987ba02960dab338
|
| 8dc7324da5bd0f2afd2ab558ab04882329a61fe8 |
|
12-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Add --include-cfi compiler option.
Decouple generation of CFI from the rest of debug symbols.
This makes it possible to generate oat with CFI but without
the rest of debug symbols.
This is in line with intention of the .eh_frame section.
The section does not have the .debug_ prefix because it
is considered somewhat different to the rest of debug symbols.
Change-Id: I32816ecd4f30ac4e0dc69d69a4993e349c737f96
|
| 2f6cdb01f74772c1c521a125776ef57ea3c73d43 |
|
11-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Relocate DWARF using .oat_patches.
The current solution is to hard-code knowledge of DWARF in the linker.
This works for simple use of DWARF, but breaks as soon as I try to do
anything more complex. Making the linker fully support DWARF would be
non-trivial task and would be essentially rewrite. Using .oat_patches
is much easier solution.
Relocating .debug_* sections required extending .oat_patches to support
more sections than just .text. I have encoded each section as
null-terminated section name followed by ULEB128 deltas.
The ULEB128 encoding shrinks .oat_patches for .text by factor of
about 6 with 64-bit compiler, and factor of 3 with 32-bit compiler.
On the other hand, it grows by the extra .oat_patches for DWARF which
were not present before (if debug symbols are included).
Overall, it is still a clear improvement even with the DWARF patches.
Change-Id: I78ffeda0f8a3da03341995a3b5ef15c954e16e9f
|
| 3b9d57ab580a0593635a7dbe3dd2e2c76274f9fa |
|
10-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Move DWARF related code in ElfWriter to its own file.
This is purely for organisation. The DWARF related code
is becoming more complex and I do not want to keep mixing
it with the ElfWriter. The ElfWriter probably should not
be dealing with the fine details of DWARF generation.
I intend to do much more work on DWARF in the future so
this situatuion would only get worse as the code grows.
The code could also use some refactoring and splitting.
However, I have mostly just moved it in this CL.
Change-Id: I6e3401a2b745ddf991b440ad7899ad7d4b921be6
|
| b536247b1ce5de640eec81dddac47802cd074363 |
|
08-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Extend the DWARF library to support .debug_info section.
Change-Id: I9916abd8db227e7a73a3311294e675be5222a709
|
| 0df9e1faed9b095b084c4eca6ee59d24fba21c9f |
|
07-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Remove duplicate of DexFile::DecodeDebugInfo.
Call DexFile::DecodeDebugInfo instead of having a local copy.
Change-Id: I4f94fd56a81fea1a2d10d5a26b9650d6d7ff9448
|
| c6b4dd8980350aaf250f0185f73e9c42ec17cd57 |
|
07-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Implement CFI for Optimizing.
CFI is necessary for stack unwinding in gdb, lldb, and libunwind.
Change-Id: I1a3480e3a4a99f48bf7e6e63c4e83a80cfee40a2
|
| 1109fb3cacc8bb667979780c2b4b12ce5bb64549 |
|
07-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Implement CFI for Quick.
CFI is necessary for stack unwinding in gdb, lldb, and libunwind.
Change-Id: Ic3b84c9dc91c4bae80e27cda02190f3274e95ae8
|
| 8c57831b2b07185ee1986b9af68a351e1ca584c3 |
|
07-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Remove the old CFI infrastructure.
Change-Id: I12a17a8a1c39ffccaa499c328ebac36e4d74dc4e
|
| 189ae8251dfb0cace95a6c21d7f8132ecb486778 |
|
04-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Proper fix for issue 20051601 - arm actually means thumb2.
Change-Id: I38e81868cbebdccec6ecbba40494612f7c1d8092
|
| 934d754a53acff144273bd7fc42f4272628bb944 |
|
02-Apr-2015 |
David Srbecky <dsrbecky@google.com> |
Fix for issue 20051601 - ARM instruction alignment.
The instructions do not have to be 4 byte aligned.
Change-Id: I58078a15190b835601094611ed01a599e2d99563
|
| 6f7158927fee233255f8e96719c374694b10cad3 |
|
30-Mar-2015 |
David Srbecky <dsrbecky@google.com> |
Write .debug_line section using the new DWARF library.
Also simplify dex to java mapping and handle mapping
in prologues and epilogues.
Change-Id: I410f06024580f2a8788f2c93fe9bca132805029a
|
| 20f85597828194c12be10d3a927999def066555e |
|
19-Mar-2015 |
Vladimir Marko <vmarko@google.com> |
Fixed layout for dex caches in boot image.
Define a fixed layout for dex cache arrays (type, method,
string and field arrays) for dex caches in the boot image.
This gives those arrays fixed offsets from the boot image
code and allows PC-relative addressing of their elements.
Use the PC-relative load on arm64 for relevant instructions,
i.e. invoke-static, invoke-direct, const-string,
const-class, check-cast and instance-of. This reduces the
arm64 boot.oat on Nexus 9 by 1.1MiB.
This CL provides the infrastructure and shows on the arm64
the gains that we can achieve by having fixed dex cache
arrays' layout. To fully use this for the boot images, we
need to implement the PC-relative addressing for other
architectures. To achieve similar gains for apps, we need
to move the dex cache arrays to a .bss section of the oat
file. These changes will be implemented in subsequent CLs.
(Also remove some compiler_driver.h dependencies to reduce
incremental build times.)
Change-Id: Ib1859fa4452d01d983fd92ae22b611f45a85d69b
|
| 5c42c29b89286e5efa4a4613132b09051ce5945b |
|
25-Feb-2015 |
Vladimir Marko <vmarko@google.com> |
Add support for .bss section in oat files.
Change-Id: I779b80b8139d9afdc28373f8c68edff5df7726ce
|
| 80b96d1a76790527f72a660ac03d9c215eed17ce |
|
19-Feb-2015 |
Vladimir Marko <vmarko@google.com> |
Replace a few std::vector with ArenaVector in Mir2Lir.
Change-Id: I7867d60afc60f57cdbbfd312f02883854d65c805
|
| e21dc3db191df04c100620965bee4617b3b24397 |
|
09-Dec-2014 |
Andreas Gampe <agampe@google.com> |
ART: Swap-space in the compiler
Introduce a swap-space and corresponding allocator to transparently
switch native allocations to memory backed by a file.
Bug: 18596910
(cherry picked from commit 62746d8d9c4400e4764f162b22bfb1a32be287a9)
Change-Id: I131448f3907115054a592af73db86d2b9257ea33
|
| 8683038c1f59bea790d8c7691e40eed7f6250e4a |
|
13-Dec-2014 |
Andreas Gampe <agampe@google.com> |
ART: Do not inline elf writer debug symbols
Using Clang, this pushes the frame size of the caller across our
limit. Thus forbid inlining. The function is only called once per
compile, impact is insignificant.
Bug: 18738594
Change-Id: I19c3f1168a5104ab508a8dbf9f2a8c035cb97e3c
|
| 807140048f82a2b87ee5bcf337f23b6a3d1d5269 |
|
21-Nov-2014 |
Mathieu Chartier <mathieuc@google.com> |
Add fast string sharpening
String sharpening changes const strings to PC relative loads instead
of always going through the dex cache. This saves code size and
probably improves performance slightly.
Before: 49602992 system@framework@boot.oat
After: 49385904 system@framework@boot.oat
Pre-cursor to removing dex_cache_strings_ field from ArtMethod.
Bug: 17643507
Change-Id: I1787f48774631eee0accafeea257aa8d0e91e8d6
|
| 6a3c1fcb4ba42ad4d5d142c17a3712a6ddd3866f |
|
31-Oct-2014 |
Ian Rogers <irogers@google.com> |
Remove -Wno-unused-parameter and -Wno-sign-promo from base cflags.
Fix associated errors about unused paramenters and implict sign conversions.
For sign conversion this was largely in the area of enums, so add ostream
operators for the effected enums and fix tools/generate-operator-out.py.
Tidy arena allocation code and arena allocated data types, rather than fixing
new and delete operators.
Remove dead code.
Change-Id: I5b433e722d2f75baacfacae4d32aef4a828bfe1b
|
| f9734550c1453f173150b12efc72b10ebf67878d |
|
27-Oct-2014 |
Ningsheng Jian <ningsheng.jian@arm.com> |
ARM: Generate $t mapping symbol to indicate thumb2 code
According to ARM ELF specification, add $t mapping symbol to oat file
to indicate that the compiled method is thumb2 code.
Change-Id: I31e4eeab0b2be7754127cc90ac40b00216b48009
|
| 0279ebb3efd653e6bb255470c99d26949c7bcd95 |
|
09-Oct-2014 |
Ian Rogers <irogers@google.com> |
Tidy ELF builder.
Don't do "if (ptr)". Use const. Use DISALLOW_COPY_AND_ASSIGN. Avoid public
member variables.
Move ValueObject to base and use in ELF builder.
Tidy VectorOutputStream to not use non-const reference arguments.
Change-Id: I2c727c3fc61769c3726de7cfb68b2d6eb4477e53
|
| f4da675bbc4615c5f854c81964cac9dd1153baea |
|
01-Aug-2014 |
Vladimir Marko <vmarko@google.com> |
Implement method calls using relative BL on ARM.
Store the linker patches with each CompiledMethod instead of
keeping them in CompilerDriver. Reorganize oat file creation
to apply the patches as we're writing the method code. Add
framework for platform-specific relative call patches in the
OatWriter. Implement relative call patches for ARM.
Change-Id: Ie2effb3d92b61ac8f356140eba09dc37d62290f8
|
| 3c5ec99e156632a734c2359642ae928102850597 |
|
24-Sep-2014 |
Andreas Gampe <agampe@google.com> |
ART: Refactor elf_writer_quick for elf section order
This writes ELF sections in approximate order (debug sections are
currently out-of-order) to avoid incompletely written files
appearing OK to ART loading code.
Bug: 17622827
Change-Id: I812fd42fcf2823dbfaf5891bf101d3b59406f263
|
| 54fc26c7350beb782d042ba61cb06284b3a367e4 |
|
05-Sep-2014 |
Andreas Gampe <agampe@google.com> |
ART: Refactor elf_writer_quick, add symbolizer
Refactors some classes in elf_writer_quick.h to elf_builder.h to
be more friendly for re-use. Use this in oatdump to add a symtab
to an oat file.
Bug: 17187621, 17322125
Change-Id: I2333291334fd98bd09cc5717fb83cb18efe3a029
|
| f9b87b1eece0e03578c4d1b627f1d5e8691a539a |
|
02-Sep-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Revert "Revert "ART: Prepare for ELF64.""
Fix was commited in "Fix frame size issue with 64-bit"
This reverts commit 7eebd95bddc7c6866ba29010d28b4f8251b43242.
Change-Id: I799b7fcab14ae24471d2f9d6d3a5e5cedd4aa7b8
|
| 7eebd95bddc7c6866ba29010d28b4f8251b43242 |
|
01-Sep-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Revert "ART: Prepare for ELF64."
Build fails on arm64 with error message:
art/compiler/elf_writer_quick.cc:1083:1: error: the frame size of 2016 bytes is larger than 1728 bytes [-Werror=frame-larger-than=]
This reverts commit 45b8f1ab2782d02830c77751f883456de799dfb6.
Change-Id: I61b2e5f27ba4b2b32eacaec51fb5e5ed12fd8066
|
| 18a49cc2f956dec648e0a13186d42374023d47a7 |
|
30-Aug-2014 |
Brian Carlstrom <bdc@google.com> |
Fix frame size issue with 64-bit
Change-Id: Ib2f33dfb79952ba39f2d7e5303a2a2e6c4b0a0f6
|
| 45b8f1ab2782d02830c77751f883456de799dfb6 |
|
27-Aug-2014 |
Tong Shen <endlessroad@google.com> |
ART: Prepare for ELF64.
Make ElfWriterQuick a template class. No functionality change.
Change-Id: I62b60dc9795c4795027e80f0bee6b54622ea49e9
|
| 33ac819bd13c6e9d42b36ae8393c37cfb1bb4cde |
|
19-Aug-2014 |
Yevgeny Rouban <yevgeny.y.rouban@intel.com> |
ART fix oat debug source map operations
Several places need to be fixed in OAT debug source map generation
(see comments in https://android-review.googlesource.com/#/c/102610/19/compiler/compiled_method.h):
1. Source Maps are deduplicated in Compiler Driver by implicit conversion
SrcMapElems to bytes. This implies incorrect operator==.
2. SrcMapElem operator < is peculiar, and cannot be applied to
SrcMapElems with negative to_ fields
3. SrcMap.Arrange method is not elegant
The fix is to introduce explicit conversion from SrcMapElem to one
signed 64-bit value, which is used as a base of two new operators < and ==.
They are correct and intuitive. DedupeHashFunc is changed to
explicitly convert array elements to byte, so the explicit type conversion
from SrcMapElem to byte is used.
Minor fix: In Line Table Programs the file index set command is generated
only if the index gets new value.
Change-Id: I5e2c03404a437254fc2db3485b22bfc1799b39b7
Signed-off-by: Yevgeny Rouban <yevgeny.y.rouban@intel.com>
|
| e3ea83811d47152c00abea24a9b420651a33b496 |
|
08-Aug-2014 |
Yevgeny Rouban <yevgeny.y.rouban@intel.com> |
ART source line debug info in OAT files
OAT files have source line information enough for ART runtime needs like
jump to/from interpreter and thread suspension. But this information
is not enough for finer grained source level debugging and low-level
profiling (VTune or perf).
This patch adds to OAT files two additional sections:
.debug_line - DWARF formatted Elf32 section with detailed source line
information (mapping from native PC to Java source lines).
In addition to the debugging symbols added using the dex2oat option
--include-debug-symbols, the source line information is added to
the section .debug_line.
The source line info can be read by many Elf reading tools like objdump,
readelf, dwarfdump, gdb, perf, VTune, ...
gdb can use this debug line information in x86. In 64-bit mode
the information can be used if the oat file is mapped in the lower
address space (address has higher 32 bits zeroed). Relocation works.
Testing:
1. art/test/run-test --host --gdb [--64] 001-HelloWorld
2. in gdb: break Main.java:19
3. in gdb: break Runtime.java:111
4. in gdb: run - stops at void java.lang.Runtime.<init>()
5. in gdb: backtrace - shows call stack down to main()
6. in gdb: continue - stops at void Main.main() (only in 32-bit mode)
7. in gdb: backtrace - shows call stack down to main()
8. objdump -W <oat-file> - addresses are from VMA range of .text
section reported by objdump -h <file>
9. dwarfdump -ka <oat-file> - no errors expected
Size of aosp-x86-eng boot.oat increased by 11% from 80.5Mb to 89.2Mb
with two sections added .debug_line (7.2Mb) and .rel.debug (1.5Mb).
Change-Id: Ib8828832686e49782a63d5529008ff4814ed9cda
Signed-off-by: Yevgeny Rouban <yevgeny.y.rouban@intel.com>
|
| 4b0d2d99b596eebd6da45679896d3af54bc5a935 |
|
06-Aug-2014 |
Alex Light <allight@google.com> |
Make patchoat use the sht_entsize when patching.
This fixes an occasional problem with multiarch use of patchoat.
Change-Id: I80799de36774720bd985704f9b709a8378bb5af5
|
| 79273802f2b788bcd3eb76edf4df1bcaa57f886f |
|
06-Aug-2014 |
Andreas Gampe <agampe@google.com> |
ART: Rework CFA frame initialization and writing code
Move eh_frame initialization code and CFI writing code to
elf_writer_quick to remove hard-wired dependencies on specific
Quick-compiler backends.
Change-Id: I27ee8ce7245da33a20c90e0086b8d4fd0a2baf4d
|
| 547cdfd21ee21e4ab9ca8692d6ef47c62ee7ea52 |
|
05-Aug-2014 |
Tong Shen <endlessroad@google.com> |
Emit CFI for x86 & x86_64 JNI compiler.
Now for host-side x86 & x86_64 ART, we are able to get complete stacktrace with even mixed C/C++ & Java stack frames.
Testing:
1. art/test/run-test --host --gdb [--64] --no-relocate 005
2. In gdb, run 'b art::Class_classForName' which is implementation of a Java native method, then 'r'
3. In gdb, run 'bt'. You should see stack frames down to main()
Change-Id: I2d17e9aa0f6d42d374b5362a15ea35a2fce96302
|
| 35e1e6ad4b50f1adbe9f93fe467766f042491896 |
|
30-Jul-2014 |
Tong Shen <endlessroad@google.com> |
1. Fix CFI for quick compiled code in x86 & x86_64;
2. Emit CFI in .eh_frame instead of .debug_frame.
With CFI, we can correctly unwind past quick generated code.
Now gdb should unwind to main() for both x86 & x86_64 host-side ART.
Note that it does not work with relocation yet.
Testing:
1. art/test/run-test --host --gdb [--64] --no-relocate 005
2. In gdb, run 'b art_quick_invoke_stub', then 'r', then 'c' a few times
3. In gdb, run 'bt'. You should see stack frames down to main()
Change-Id: I5350d4097dc3d360a60cb17c94f1d02b99bc58bb
|
| 53cb16b98acf3cf6f3a1e2204ad4958ecf1b5a3c |
|
12-Jun-2014 |
Alex Light <allight@google.com> |
Add patchoat tool to Art.
Add a new executable called patchoat to art. This tool takes already
compiled images and oat files and changes their base address, acting as
a cheap form of relocation.
Add a --include-patch-information flag to dex2oat and code to add
required patch information to oat files created with the quick compiler.
Bug: 15358152
Change-Id: Ie0c580db45bb14ec180deb84930def6c3628d97d
|
| eefbe39799126afdf7d315a79359b9da23d1cef5 |
|
08-Jul-2014 |
Alex Light <allight@google.com> |
Fix some style nitpicks
Change-Id: Icfdd327f4ddf129f0a8607162c09ba271c1d49d9
|
| 3470ab4011b5e18d590d5375e2f13a1e3bd69222 |
|
18-Jun-2014 |
Alex Light <allight@google.com> |
Fixed gdb support and added some ElfFile functions
Fixed gdb support so that it would continue working even when debug
symbols or other sections are included in the elf file. Also made it
actually read parts of the DWARF information so it should work even if
there are minor changes to how and where DWARF information is written
out.
Added a dwarf.h file with the dwarf constants.
Added a FindSectionByName function, a FindDynamicSymbol function, and
the ability to specify the mmap protection and flags directly if we are
mapping in the whole file.
Modified elf_writer_quick.cc to use the dwarf constants from dwarf.h.
Change-Id: I09e15c425fab252b331a2e4719863552e8b6b137
|
| 78382fa44ee505cf16835e4d22515e7252a90864 |
|
07-Jun-2014 |
Alex Light <allight@google.com> |
Optionally add debug symbols to ELF files made by quick compiler.
Added debug symbols to ELF files created by dex2oat using
the quick compiler. Adds two flags --include-debug-symbols and
--no-include-debug-symbols for dex2oat that control the inclusion of these
debug symbols. Debug info is added by default if kIsDebugBuild is true.
Fixed bug where Intel DWARF information would not correctly deal with
deduplicated code the binary.
Changed the portable compiler code path in dex2oat.cc so that symbols
will not be stripped when run with --include-debug-symbols.
Change-Id: Ia2eb2f654dedf0e5e8606f7744e05b8d14155fb1
|
| 8758c64a40e74ebb492a348556ec7b25a89c1491 |
|
11-Jun-2014 |
Brian Carlstrom <bdc@google.com> |
Fix build by converting from bionic to elfutils version of elf.h
Change-Id: I70a3b6fedae6d290a4f01d4ae8fdd726b8aaa90f
|
| b12f34742be4adaa804cc0d388ba51603bb95955 |
|
11-Jun-2014 |
Brian Carlstrom <bdc@google.com> |
Revert "Revert "Rewrite ElfWriterQuick to make it more modular.""
This reverts commit 35f72251e722cad03a08e8ceacbd5c244eab5c29.
|
| 35f72251e722cad03a08e8ceacbd5c244eab5c29 |
|
11-Jun-2014 |
Brian Carlstrom <bdc@google.com> |
Revert "Rewrite ElfWriterQuick to make it more modular."
This reverts commit ad13cb2db8c4a64298ed85c82ae7366ee73f7640.
Change-Id: Iee1d1fd08a008d7d382c181cb352ef46a6eaf5d3
|
| ad13cb2db8c4a64298ed85c82ae7366ee73f7640 |
|
05-Jun-2014 |
Alex Light <allight@google.com> |
Rewrite ElfWriterQuick to make it more modular.
Remade ElfWriterQuick so that it would use a more modular design and allow
the addition of additional features onto it in a more straightforward way.
Change-Id: I07971eb2080890645fab1b02c7305697e85d3a64
|
| ffddfdf6fec0b9d98a692e27242eecb15af5ead2 |
|
03-Jun-2014 |
Tim Murray <timmurray@google.com> |
DO NOT MERGE
Merge ART from AOSP to lmp-preview-dev.
Change-Id: I0f578733a4b8756fd780d4a052ad69b746f687a9
|
| 57ef2d33b369ced7eaf95b996bba473702f636cf |
|
30-May-2014 |
Alex Light <allight@google.com> |
Moved a BufferedOutputStream onto the heap to reduce stack usage.
Bug: 15278350
Change-Id: I6aa7cf3979453e555686610e90520b762b812f8c
|
| 53bee42d0f8b454ff4fe5b7e42bdb070c786e3eb |
|
05-Apr-2014 |
Mathieu Chartier <mathieuc@google.com> |
Fix using kArm instead of kThumb2.
Change-Id: I0f3bf93b38d910bc57d017693675829b6a39bc18
|
| b95a5345ae4217b70ca36f0cced92f68dda7caf5 |
|
12-Mar-2014 |
Stuart Monteith <stuart.monteith@arm.com> |
AArch64: Add arm64 runtime support.
Adds support for arm64 to ART. Assembler stubs are sufficient for
down calls into interpreter. JNI compiler and generics are not finished.
Basic Generic JNI functionality.
Change-Id: I4a07c79d1e037b9f5746673480e32cf456867b82
|
| befbd5731ecca08f08780ee28a913d08ffb14656 |
|
06-Mar-2014 |
Ian Rogers <irogers@google.com> |
Fix host architecture for 64bit.
Also, hack x86 assembler for use as a x86-64 trampoline compiler's assembler.
Implement missing x86-64 quick resolution trampoline.
Add x86-64 to the quick elf writer.
Change-Id: I08216c67014a83492ada12898ab8000218ba7bb4
|
| ae9fd93c39a341e2dffe15c61cc7d9e841fa92c4 |
|
11-Feb-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Tell GDB about Quick ART generated code
This is actually a lot of work. To do this, we need:
.debug_info
.debug_abbrev
.debug_frame
.debug_str
These are generated into the OAT file by OatWriter and ElfWriterQuick.
Since the Quick ART runtime doesn't use dlopen to load the OAT files,
GDB can't find this information. Use the alternate GDB JIT interface,
which can be invoked at runtime. To use this interface, an ELF image
needs to be built in memory. Read the information from the OAT file,
fixup the addresses to point to the real locations, add a symbol table
to hold the .text symbol, and then let GDB know about the information,
which will be read from the runtime address space.
This is quite primitive now, and could be cleaned up considerably. It
probably needs symbol table entries for the methods, and descriptions of
parameters and return types.
Currently only supported for X86.
This defaults to enabled for debug builds. Added dexoat --gen-gdb-info
and --no-gen-gdb-info flags to override.
Change-Id: I4d18b2370f6dfaa00c8cc1925f10717be3bd1a62
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 3d504075f7c1204d581923460754bf6d3714b13f |
|
01-Mar-2014 |
Ian Rogers <irogers@google.com> |
Make out arguments non-reference types.
Also, tidy some portable related code.
Change-Id: I67c8aa52eef8b556ca117ecda1b1e75465ba06a5
|
| 50cfe74daaece80853cb3b45d4338329b7d0345b |
|
19-Feb-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Remove dependency on llvm/Support/ELF.h by using linux's elf.h.
Change-Id: Iefe66af9958641ac7f08fdc22f438d976e5b4d54
|
| c6dfdacea2fd9e268f70328805b0366cdd6b7b9e |
|
27-Aug-2013 |
Brian Carlstrom <bdc@google.com> |
Add buffering to ELF file generation
Bug: 10496017
Change-Id: I3cbad249e0fb33f726bd0a504b3b6bd9b4f759c8
|
| e2be9da597c69b92ffb707f2101b96076578b5e6 |
|
26-Aug-2013 |
Mathieu Chartier <mathieuc@google.com> |
Fix typo in elf writer.
We were writing the wrong number of characters to the elf file.
Caught by valgrind.
Change-Id: Iaf5a51bc7f6dd23cc5e68992dc033610fd0c5c79
|
| 8718359763aa3dcd0033a2d2e67e0cb145fb9c53 |
|
14-Aug-2013 |
Anwar Ghuloum <anwarg@google.com> |
Add -XX:+DisableExplicitGC option for Art
Phone boots. Also removed elfwriter spam from logcat.
Change-Id: I09b9d2fc40ebfeb6f1c2f12153b0ad4134bb5159
|
| 7934ac288acfb2552bb0b06ec1f61e5820d924a4 |
|
26-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix cpplint whitespace/comments issues
Change-Id: Iae286862c85fb8fd8901eae1204cd6d271d69496
|
| c50d8e11a098cc5c6239aa86b47d4fcf8cbb4899 |
|
24-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Remove OatWriter buffering to memory for ElfWriterQuick
This allows the oat contents to be directly written to the file.
Change-Id: Ibc7ddf57477b152f07784b52f7334be73fd22833
|
| 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
|