| 4e67841e99e4a206133e7010653ccd132682296a |
|
09-Sep-2014 |
Mathieu Chartier <mathieuc@google.com> |
Change Reference.get() intrinsic to Reference.getReferent().
The reference intrinsic was incorrectly inlining
PhantomReference.get(). We now get around this by adding a
layer of indirection. Reference.get() now calls getReferent()
which is intrinsified and inlined.
Requires:
https://android-review.googlesource.com/#/c/107100/
Bug: 17429865
(cherry picked from commit cd48f2d86197d4fe87cc88077bc4af5ba66e5295)
Change-Id: Ie91e70abf43cedf3c707c7bb8a5059e19d2a2577
|
| 8c18c2aaedb171f9b03ec49c94b0e33449dc411b |
|
06-Aug-2014 |
Andreas Gampe <agampe@google.com> |
ART: Generate chained compare-and-branch for short switches
Refactor Mir2Lir to generate chained compare-and-branch sequences
for short switches on all architectures.
Bug: 16241558
(cherry picked from commit 48971b3242e5126bcd800cc9c68df64596b43d13)
Change-Id: I0bb3071b8676523e90e0258e9b0e3fd69c1237f4
|
| c76c614d681d187d815760eb909e5faf488a3c35 |
|
05-Aug-2014 |
Andreas Gampe <agampe@google.com> |
ART: Refactor long ops in quick compiler
Make GenArithOpLong virtual. Let the implementation in gen_common be
very basic, without instruction-set checks, and meant as a fall-back.
Backends should implement and dispatch to code for better implementations.
This allows to remove the GenXXXLong virtual methods from Mir2Lir, and
clean up the backends (especially removing some LOG(FATAL) implementations).
Change-Id: I6366443c0c325c1999582d281608b4fa229343cf
|
| 8081d2b8d7a743729557051d0294e040e61c747a |
|
31-Jul-2014 |
Vladimir Marko <vmarko@google.com> |
Create allocator adapter for using Arena in std containers.
Create ArenaAllocatorAdapter, similar to the existing
ScopedArenaAllocatorAdapter, for allocating memory for
standard containers via the ArenaAllocator. Add the ability
to specify allocation kind rather than just kArenaAllocSTL
to both adapters. Move the scoped arena allocator to the
scoped_arena_containers.h header file.
Define template aliases for containers using the new adapter
and change a few MIRGraph and Mir2Lir members to use them.
Change-Id: I9bbc50248e0fed81729497b848cb29bf68444268
|
| c763e350da562b0c6bebf10599588d4901140e45 |
|
04-Jul-2014 |
Matteo Franchin <matteo.franchin@arm.com> |
AArch64: Implement InexpensiveConstant methods.
Implement IsInexpensiveConstant and friends for A64.
Also extending the methods to take the opcode with respect to which
the constant is inexpensive. Additionally, logical operations (i.e.
and, or, xor) can now handle the immediates 0 and ~0 (which are not
logical immediates).
Change-Id: I46ce1287703765c5ab54983d13c1b3a1f5838622
|
| 6bbf0967d217ab2b7bdbb78bfd076b8fb07a44e8 |
|
14-Jul-2014 |
Alexei Zavjalov <alexei.zavjalov@intel.com> |
ART: Implement the easy long division/remainder by a constant
Also optimizes long/int divisions by power-of-two values.
Also do some clean-up.
Change-Id: Ie414e64aac251c81361ae107d157c14439e6dab5
Signed-off-by: Alexei Zavjalov <alexei.zavjalov@intel.com>
|
| 2eba1fa7e9e5f91e18ae3778d529520bd2c78d55 |
|
31-Jul-2014 |
Serban Constantinescu <serban.constantinescu@arm.com> |
AArch64: Add inlining support for ceil(), floor(), rint(), round()
This patch adds inlining support for the following Math, StrictMath
methods in the ARM64 backend:
* double ceil(double)
* double floor(double)
* double rint(double)
* long round(double)
* int round(float)
Also some cleanup.
Change-Id: I9f5a2f4065b1313649f4b0c4380b8176703c3fe1
Signed-off-by: Serban Constantinescu <serban.constantinescu@arm.com>
|
| 63999683329612292d534e6be09dbde9480f1250 |
|
15-Jul-2014 |
Serban Constantinescu <serban.constantinescu@arm.com> |
Revert "Revert "Enable Load Store Elimination for ARM and ARM64""
This patch refactors the implementation of the LoadStoreElimination
optimisation pass. Please note that this pass was disabled and not
functional for any of the backends.
The current implementation tracks aliases and handles DalvikRegs as well
as Heap memory regions. It has been tested and it is known to optimise
out the following:
* Load - Load
* Store - Load
* Store - Store
* Load Literals
Change-Id: I3aadb12a787164146a95bc314e85fa73ad91e12b
|
| c32447bcc8c36ee8ff265ed678c7df86936a9ebe |
|
27-Jul-2014 |
Bill Buzbee <buzbee@android.com> |
Revert "Enable Load Store Elimination for ARM and ARM64"
On extended testing, I'm seeing a CHECK failure at utility_arm.cc:1201.
This reverts commit fcc36ba2a2b8fd10e6eebd21ecb6329606443ded.
Change-Id: Icae3d49cd7c8fcab09f2f989cbcb1d7e5c6d137a
|
| fcc36ba2a2b8fd10e6eebd21ecb6329606443ded |
|
15-Jul-2014 |
Serban Constantinescu <serban.constantinescu@arm.com> |
Enable Load Store Elimination for ARM and ARM64
This patch refactors the implementation of the LoadStoreElimination
optimisation pass. Please note that this pass was disabled and not
functional for any of the backends.
The current implementation tracks aliases and handles DalvikRegs as well
as Heap memory regions. It has been tested and it is known to optimise
out the following:
* Load - Load
* Store - Load
* Store - Store
* Load Literals
Change-Id: Iefae9b696f87f833ef35c451ed4d49c5a1b6fde0
|
| 984305917bf57b3f8d92965e4715a0370cc5bcfb |
|
28-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Rework quick entrypoint code in Mir2Lir, cleanup
To reduce the complexity of calling trampolines in generic code,
introduce an enumeration for entrypoints. Introduce a header that lists
the entrypoint enum and exposes a templatized method that translates an
enum value to the corresponding thread offset value.
Call helpers are rewritten to have an enum parameter instead of the
thread offset. Also rewrite LoadHelper and GenConversionCall this way.
It is now LoadHelper's duty to select the right thread offset size.
Introduce InvokeTrampoline virtual method to Mir2Lir. This allows to
further simplify the call helpers, as well as make OpThreadMem specific
to X86 only (removed from Mir2Lir).
Make GenInlinedCharAt virtual, move a copy to X86 backend, and simplify
both copies. Remove LoadBaseIndexedDisp and OpRegMem from Mir2Lir, as they
are now specific to X86 only.
Remove StoreBaseIndexedDisp from Mir2Lir, as it was only ever used in the
X86 backend.
Remove OpTlsCmp from Mir2Lir, as it was only ever used in the X86 backend.
Remove OpLea from Mir2Lir, as it was only ever defined in the X86 backend.
Remove GenImmedCheck from Mir2Lir as it was neither used nor implemented.
Change-Id: If0a6182288c5d57653e3979bf547840a4c47626e
|
| bebee4fd10e5db6cb07f59bc0f73297c900ea5f0 |
|
16-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Refactor GenSelect, refactor gen_common accordingly
This adds a GenSelect method meant for selection of constants. The
general-purpose GenInstanceof code is refactored to take advantage of
this. This cleans up code and squashes a branch-over on ARM64 to a
cset.
Also add a slow-path for type initialization in GenInstanceof.
Bug: 16241558
(cherry picked from commit 90969af6deb19b1dbe356d62fe68d8f5698d3d8f)
Change-Id: Ie4494858bb8c26d386cf2e628172b81bba911ae5
|
| 0f45f22eb3c52f0ece4c56989180e79c6680d825 |
|
15-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Throw StackOverflowError in native code
Initialize stack-overflow errors in native code to be able to reduce
the preserved area size of the stack.
Includes a refactoring away from constexpr in instruction_set.h to allow
for easy changing of the values.
Bug: 16256184
(cherry picked from commit 7ea6f79bbddd69d5db86a8656a31aaaf64ae2582)
Change-Id: I117cc8485f43da5f0a470f0f5e5b3dc3b5a06246
|
| 9ee4519afd97121f893f82d41d23164fc6c9ed34 |
|
17-Jul-2014 |
Serguei Katkov <serguei.i.katkov@intel.com> |
x86: GenSelect utility update
The is follow-up https://android-review.googlesource.com/#/c/101396/
to make x86 GenSelectConst32 implementation complete.
Change-Id: I69f318e18093f9a5b00f8f00f0f1c2e4ff7a9ab2
Signed-off-by: Serguei Katkov <serguei.i.katkov@intel.com>
|
| f9d6aede77c700118e225f8312cd888262b77862 |
|
17-Jul-2014 |
Vladimir Marko <vmarko@google.com> |
Use vabs/fabs on arm/arm64 for intrinsic abs().
Bug: 11579369
(cherry picked from 5030d3ee8c6fe10394912ede107cbc8df63b7b16)
Change-Id: I7b0596a8e7e3c87a93b225519c5aeedfe4f22e6d
|
| 7ea6f79bbddd69d5db86a8656a31aaaf64ae2582 |
|
15-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Throw StackOverflowError in native code
Initialize stack-overflow errors in native code to be able to reduce
the preserved area size of the stack.
Includes a refactoring away from constexpr in instruction_set.h to allow
for easy changing of the values.
Change-Id: I117cc8485f43da5f0a470f0f5e5b3dc3b5a06246
|
| 147eb41b53729ec8d5c188d1cac90964a51afb8a |
|
11-Jul-2014 |
Dave Allison <dallison@google.com> |
Revert "Revert "Revert "Revert "Add implicit null and stack checks for x86""""
This reverts commit 0025a86411145eb7cd4971f9234fc21c7b4aced1.
Bug: 16256184
Change-Id: Ie0760a0c293aa3b62e2885398a8c512b7a946a73
Conflicts:
compiler/dex/quick/arm64/target_arm64.cc
compiler/image_test.cc
runtime/fault_handler.cc
|
| f12feb8e0e857f2832545b3f28d31bad5a9d3903 |
|
17-Jul-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Stack overflow checks and NPE checks for optimizing.
Change-Id: I59e97448bf29778769b79b51ee4ea43f43493d96
|
| 5030d3ee8c6fe10394912ede107cbc8df63b7b16 |
|
17-Jul-2014 |
Vladimir Marko <vmarko@google.com> |
Use vabs/fabs on arm/arm64 for intrinsic abs().
Bug: 11579369
Change-Id: If09da85e22786faa13a2d74f62cee68ea67bd087
|
| d85614222fa062ec809af9d65f04ab6b7dc1c248 |
|
11-Jul-2014 |
Fred Shih <ffred@google.com> |
Revert "Revert "Revert "Revert "Add intrinsic for Reference.get()""""
Fixed TargetReg issue causing build failure for x86.
This reverts commit 9e82bd3f0ce9e5f5777bea2f752ff3e251d32f9f.
(cherry picked from commit 4ee7a665e7f9cd2c5ace2d6304e33f64067b209f)
Change-Id: I555f4e06955711262e6b37ffbeabee9698ec695c
|
| 90969af6deb19b1dbe356d62fe68d8f5698d3d8f |
|
16-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Refactor GenSelect, refactor gen_common accordingly
This adds a GenSelect method meant for selection of constants. The
general-purpose GenInstanceof code is refactored to take advantage of
this. This cleans up code and squashes a branch-over on ARM64 to a
cset.
Also add a slow-path for type initialization in GenInstanceof.
Change-Id: Ie4494858bb8c26d386cf2e628172b81bba911ae5
|
| 69dfe51b684dd9d510dbcb63295fe180f998efde |
|
11-Jul-2014 |
Dave Allison <dallison@google.com> |
Revert "Revert "Revert "Revert "Add implicit null and stack checks for x86""""
This reverts commit 0025a86411145eb7cd4971f9234fc21c7b4aced1.
Bug: 16256184
Change-Id: Ie0760a0c293aa3b62e2885398a8c512b7a946a73
|
| d9cb8ae2ed78f957a773af61759432d7a7bf78af |
|
09-Jul-2014 |
Douglas Leung <douglas@mips.com> |
Fix art test failures for Mips.
This patch fixes the following art test failures for Mips:
003-omnibus-opcodes
030-bad-finalizer
041-narrowing
059-finalizer-throw
Change-Id: I4e0e9ff75f949c92059dd6b8d579450dc15f4467
Signed-off-by: Douglas Leung <douglas@mips.com>
|
| 4ee7a665e7f9cd2c5ace2d6304e33f64067b209f |
|
11-Jul-2014 |
Fred Shih <ffred@google.com> |
Revert "Revert "Revert "Revert "Add intrinsic for Reference.get()""""
Fixed TargetReg issue causing build failure for x86.
This reverts commit 9e82bd3f0ce9e5f5777bea2f752ff3e251d32f9f.
Change-Id: I7e6a526954467aaf68deeed999880dfe9aa5f06e
|
| ed7a0f2fb84b200ab6ef34e30dcbba4c0cf8d435 |
|
10-Jun-2014 |
Matteo Franchin <matteo.franchin@arm.com> |
AArch64: improve usage of TargetReg() and friends.
TargetReg(arg1) does now always return a 32-bit register. We also avoid
using this function directly and rather use the two-arguments overload
or TargetPtrReg().
Change-Id: I746b3c29a2a2553b399b5c3e7ee3887c7e7c52c3
|
| af263df7f643e699abf622c64447d31bacc14c34 |
|
12-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Change GenPCUseDefEncoding(), turn on Load Hoisting for ARM64
This defines the PC resource mask as empty, as the PC is not
accessible on ARM64.
Unify code paths with x86 in LoadStoreElimination and LoadHoisting.
Change-Id: Iea8b9e666f306c7a6ff52b6c5bf7e05b35346b2c
|
| a9b870b73a155ce70c867d5b3f9758fab0b45f07 |
|
11-Jul-2014 |
Christopher Ferris <cferris@google.com> |
Revert "Add intrinsic for Reference.get()"
This reverts commit 460503b13bc894828a2d2d47d09e5534b3e91aa1.
Change-Id: Ie63f43049307e02e3b90f4e034abc9ea54ca4e24
|
| ccc60264229ac96d798528d2cb7dbbdd0deca993 |
|
05-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Rework TargetReg(symbolic_reg, wide)
Make the standard implementation in Mir2Lir and the specialized one
in the x86 backend return a pair when wide = "true". Introduce
WideKind enumeration to improve code readability. Simplify generic
code based on this implementation.
Change-Id: I670d45aa2572eedfdc77ac763e6486c83f8e26b4
|
| 59a42afc2b23d2e241a7e301e2cd68a94fba51e5 |
|
04-Jul-2014 |
Serguei Katkov <serguei.i.katkov@intel.com> |
Update counting VR for promotion
For 64-bit it makes sense to compute VR uses together for
int and long because core reg is shared.
Change-Id: Ie8676ece12c928d090da2465dfb4de4e91411920
Signed-off-by: Serguei Katkov <serguei.i.katkov@intel.com>
|
| e9f3e71c90094e87ff83bd5449a2fc4d65f717b2 |
|
04-Jul-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Updates to help classes derived from X86Mir2Lir
Just a couple of extra changes to help me out. These changes won't
affect anyone else.
Change-Id: I0e0985a4f16822d5cbfabbf81c9902d34ebdb5da
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 7fb36ded9cd5b1d254b63b3091f35c1e6471b90e |
|
10-Jul-2014 |
Dave Allison <dallison@google.com> |
Revert "Revert "Add implicit null and stack checks for x86""
Fixes x86_64 cross compile issue. Removes command line options
and property to set implicit checks - this is hard coded now.
This reverts commit 3d14eb620716e92c21c4d2c2d11a95be53319791.
Change-Id: I5404473b5aaf1a9c68b7181f5952cb174d93a90d
|
| d4415e8bd04c4a9367744ff0149597b4f37a0e0a |
|
11-Jul-2014 |
Christopher Ferris <cferris@google.com> |
Revert "Revert "Add intrinsic for Reference.get()""
This reverts commit a9b870b73a155ce70c867d5b3f9758fab0b45f07.
Change-Id: Ic2a9b47f2b911bef4b764d10bc33cf000e4b4211
|
| 9e82bd3f0ce9e5f5777bea2f752ff3e251d32f9f |
|
11-Jul-2014 |
Sebastien Hertz <shertz@google.com> |
Revert "Revert "Revert "Add intrinsic for Reference.get()"""
This reverts commit d4415e8bd04c4a9367744ff0149597b4f37a0e0a.
Change-Id: I34553ccbdcfea35c7742d21be2a74dc7085ab2a0
|
| 0025a86411145eb7cd4971f9234fc21c7b4aced1 |
|
11-Jul-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Revert "Revert "Revert "Add implicit null and stack checks for x86"""
Broke the build.
This reverts commit 7fb36ded9cd5b1d254b63b3091f35c1e6471b90e.
Change-Id: I9df0e7446ff0913a0e1276a558b2ccf6c8f4c949
|
| 460503b13bc894828a2d2d47d09e5534b3e91aa1 |
|
18-Jun-2014 |
Fred Shih <ffred@google.com> |
Add intrinsic for Reference.get()
Added an intrinsic function for Reference.get(). Return immediately
without going through JNI if the slow path is not currently in use.
Otherwise, branch off to the the existing JNI function.
Approximately 47x speedup for cases where slow path is not enabled.
Change-Id: I13ad65a356fe4e104d8d83980694dc2740d7d039
|
| 34e826ccc80dc1cf7c4c045de6b7f8360d504ccf |
|
29-May-2014 |
Dave Allison <dallison@google.com> |
Add implicit null and stack checks for x86
This adds compiler and runtime changes for x86
implicit checks. 32 bit only.
Both host and target are supported.
By default, on the host, the implicit checks are null pointer and
stack overflow. Suspend is implemented but not switched on.
Change-Id: I88a609e98d6bf32f283eaa4e6ec8bbf8dc1df78a
|
| 3d14eb620716e92c21c4d2c2d11a95be53319791 |
|
10-Jul-2014 |
Dave Allison <dallison@google.com> |
Revert "Add implicit null and stack checks for x86"
It breaks cross compilation with x86_64.
This reverts commit 34e826ccc80dc1cf7c4c045de6b7f8360d504ccf.
Change-Id: I34ba07821fc0a022fda33a7ae21850957bbec5e7
|
| 70c4f06f9965cdb9319a2c85f65acda20086d765 |
|
25-Jun-2014 |
DaniilSokolov <daniil.y.sokolov@intel.com> |
ART: Intrinsic implementation for java.lang.System.arraycopy.
Implements intrinsic for java.lang.System.arraycopy(char[], int, char[], int, int) -
this method is internal to android class libraries and used in such classes as StringBuffer and
StringBuilder. It is not possible to call it from application code. The intrinsic for
this method is implemented as inline method (assembly code is generated manually).
The intrinsic is x86 32 bit only.
Change-Id: Id1b1e0a20d5f6d5f5ebfe1fdc2447b6d8a515432
Signed-off-by: Daniil Sokolov <daniil.y.sokolov@intel.com>
|
| a77ee5103532abb197f492c14a9e6fb437054e2a |
|
02-Jul-2014 |
Chao-ying Fu <chao-ying.fu@intel.com> |
x86_64: TargetReg update for x86
Also includes changes in common code. Elimination of use of TargetReg
with one parameter and direct access to special target registers.
Change-Id: Ied2c1f87d4d1e4345248afe74bca40487a46a371
Signed-off-by: Serguei Katkov <serguei.i.katkov@intel.com>
Signed-off-by: Chao-ying Fu <chao-ying.fu@intel.com>
|
| b5860fb459f1ed71f39d8a87b45bee6727d79fe8 |
|
22-Jun-2014 |
buzbee <buzbee@google.com> |
Register promotion support for 64-bit targets
Not sufficiently tested for 64-bit targets, but should be
fairly close.
A significant amount of refactoring could stil be done, (in
later CLs).
With this change we are not making any changes to the vmap
scheme. As a result, it is a requirement that if a vreg
is promoted to both a 32-bit view and the low half of a
64-bit view it must share the same physical register. We
may change this restriction later on to allow for more flexibility
for 32-bit Arm.
For example, if v4, v5, v4/v5 and v5/v6 are all hot enough to
promote, we'd end up with something like:
v4 (as an int) -> r10
v4/v5 (as a long) -> r10
v5 (as an int) -> r11
v5/v6 (as a long) -> r11
Fix a couple of ARM64 bugs on the way...
Change-Id: I6a152b9c164d9f1a053622266e165428045362f3
|
| 255e014542b2180620230e4d9d6000ae06846bbd |
|
04-Jul-2014 |
Matteo Franchin <matteo.franchin@arm.com> |
Aarch64: fix references handling in Load*Indexed.
Fix the way we handle references in Load/StoreBaseIndexed and friends.
We assume references are 64-bit RegStorage entities, with the
difference that they are load as 32-bit values.
Change-Id: I7fe987ef9e97e9a5042b85378b33d1e85710d8b5
|
| 23abec955e2e733999a1e2c30e4e384e46e5dde4 |
|
02-Jul-2014 |
Serban Constantinescu <serban.constantinescu@arm.com> |
AArch64: Add few more inline functions
This patch adds inlining support for the following functions:
* Math.max/min(long, long)
* Math.max/min(float, float)
* Math.max/min(double, double)
* Integer.reverse(int)
* Long.reverse(long)
Change-Id: Ia2b1619fd052358b3a0d23e5fcbfdb823d2029b9
Signed-off-by: Serban Constantinescu <serban.constantinescu@arm.com>
|
| dd64450b37776f68b9bfc47f8d9a88bc72c95727 |
|
01-Jul-2014 |
Elena Sayapina <elena.v.sayapina@intel.com> |
x86_64: Unify 64-bit check in x86 compiler
Update x86-specific Gen64Bit() check with the CompilationUnit target64 field
which is set using unified Is64BitInstructionSet(InstructionSet) check.
Change-Id: Ic00ac863ed19e4543d7ea878d6c6c76d0bd85ce8
Signed-off-by: Elena Sayapina <elena.v.sayapina@intel.com>
|
| 4b537a851b686402513a7c4a4e60f5457bb8d7c1 |
|
01-Jul-2014 |
Andreas Gampe <agampe@google.com> |
ART: Quick compiler: More size checks, add TargetReg variants
Add variants for TargetReg for requesting specific register usage,
e.g., wide and ref. More register size checks.
With code adapted from https://android-review.googlesource.com/#/c/98605/.
Change-Id: I852d3be509d4dcd242c7283da702a2a76357278d
|
| de68676b24f61a55adc0b22fe828f036a5925c41 |
|
24-Jun-2014 |
Andreas Gampe <agampe@google.com> |
Revert "ART: Split out more cases of Load/StoreRef, volatile as parameter"
This reverts commit 2689fbad6b5ec1ae8f8c8791a80c6fd3cf24144d.
Breaks the build.
Change-Id: I9faad4e9a83b32f5f38b2ef95d6f9a33345efa33
|
| 3c12c512faf6837844d5465b23b9410889e5eb11 |
|
24-Jun-2014 |
Andreas Gampe <agampe@google.com> |
Revert "Revert "ART: Split out more cases of Load/StoreRef, volatile as parameter""
This reverts commit de68676b24f61a55adc0b22fe828f036a5925c41.
Fixes an API comment, and differentiates between inserting and appending.
Change-Id: I0e9a21bb1d25766e3cbd802d8b48633ae251a6bf
|
| 2689fbad6b5ec1ae8f8c8791a80c6fd3cf24144d |
|
23-Jun-2014 |
Andreas Gampe <agampe@google.com> |
ART: Split out more cases of Load/StoreRef, volatile as parameter
Splits out more cases of ref registers being loaded or stored. For
code clarity, adds volatile as a flag parameter instead of a separate
method.
On ARM64, continue cleanup. Add flags to print/fatal on size mismatches.
Change-Id: I30ed88433a6b4ff5399aefffe44c14a5e6f4ca4e
|
| c61b3c984c509d5f7c8eb71b853c81a34b5c28ef |
|
18-Jun-2014 |
Matteo Franchin <matteo.franchin@arm.com> |
AArch64: implement easy division and reminder.
This implements easy division and reminder for integer only (32-bit).
The optimisation applies to div/rem by powers of 2 and to div by small
literals (between 3-15).
Change-Id: I71be7c4de5d2e2e738b88984f13efb08f4388a19
|
| 7cd26f355ba83be75b72ed628ed5ee84a3245c4f |
|
19-Jun-2014 |
Andreas Gampe <agampe@google.com> |
ART: Target-dependent stack overflow, less check elision
Refactor the separate stack overflow reserved sizes from thread.h
into instruction_set.h and make sure they're used in the compiler.
Refactor the decision on when to elide stack overflow checks:
especially with large interpreter stack frames, it is not a good
idea to elide checks when the frame size is even close to the
reserved size. Currently enforce checks when the frame size is
>= 2KB, but make sure that frame sizes 1KB and below will elide
the checks (number from experience).
Bug: 15728765
Change-Id: I016bfd3d8218170cbccbd123ed5e2203db167c06
|
| 7071c8d5885175a746723a3b38a347855965be08 |
|
05-Mar-2014 |
Yixin Shou <yixin.shou@intel.com> |
Add x86 inlined abs method for float/double
Add the optimized implementation of inlined abs method for
float/double for X86 side.
Change-Id: I2f367542f321d88a976129f9f7156fd3c2965c8a
Signed-off-by: Yixin Shou <yixin.shou@intel.com>
|
| 4c115b85cc48f4dfc8fc2b0484ddfeb29f02d658 |
|
17-Jun-2014 |
Vladimir Marko <vmarko@google.com> |
Revert "Add x86 inlined abs method for float/double"
This reverts commit e88b89ad1d1a583daf205c7a387ba13f549f95f1.
Change-Id: I2ba21b7442ba3696482d45001e6bd32e8baf9d1f
|
| e88b89ad1d1a583daf205c7a387ba13f549f95f1 |
|
05-Mar-2014 |
Yixin Shou <yixin.shou@intel.com> |
Add x86 inlined abs method for float/double
Add the optimized implementation of inlined abs method for
float/double for X86 side.
Change-Id: I4e095644a90524354040174954c1e127c7bb4ee2
Signed-off-by: Yixin Shou <yixin.shou@intel.com>
|
| 5aa6e04061ced68cca8111af1e9c19781b8a9c5d |
|
14-Jun-2014 |
Ian Rogers <irogers@google.com> |
Tidy x86 assembler.
Use helper functions to compute when the kind has a SIB, a ModRM and RegReg
form.
Change-Id: I86a5cb944eec62451c63281265e6974cd7a08e07
|
| 169489b4f4be8c5dd880ba6f152948324d22ff79 |
|
11-Jun-2014 |
Serban Constantinescu <serban.constantinescu@arm.com> |
AArch64: Add support for inlined methods
This patch adds support for Arm64 inlined methods.
Change-Id: Ic6aeed6d2d32f65cd1e63cf482f83cdcf958798a
|
| 8dea81ca9c0201ceaa88086b927a5838a06a3e69 |
|
06-Jun-2014 |
Vladimir Marko <vmarko@google.com> |
Rewrite use/def masks to support 128 bits.
Reduce LIR memory usage by holding masks by pointers in the
LIR rather than directly and using pre-defined const masks
for the common cases, allocating very few on the arena.
Change-Id: I0f6d27ef6867acd157184c8c74f9612cebfe6c16
|
| 58994cdb00b323339bd83828eddc53976048006f |
|
16-May-2014 |
Dmitry Petrochenko <dmitry.petrochenko@intel.com> |
x86_64: Hard Float ABI support in QCG
This patch shows our efforts on resolving the ART limitations:
- passing "float"/"double" arguments via FPR
- passing "long" arguments via single GPR, not pair
- passing more than 3 agruments via GPR.
Work done:
- Extended SpecialTargetRegister enum with kARG4, kARG5, fARG4..fARG7.
- Created initial LoadArgRegs/GenDalvikX/FlushIns version in X86Mir2Lir.
- Unlimited number of long/double/float arguments support
- Refactored (v2)
Change-Id: I5deadd320b4341d5b2f50ba6fa4a98031abc3902
Signed-off-by: Serguei Katkov <serguei.i.katkov@intel.com>
Signed-off-by: Dmitry Petrochenko <dmitry.petrochenko@intel.com>
Signed-off-by: Chao-ying Fu <chao-ying.fu@intel.com>
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 089142cf1d0c028b5a7c703baf0b97f4a4ada3f7 |
|
05-Jun-2014 |
Vladimir Marko <vmarko@google.com> |
Avoid register pool allocations on the heap.
Create a helper template class ArrayRef and use it instead
of std::vector<> for register pools in target_<arch>.cc to
avoid these heap allocations during program startup.
Change-Id: I4ab0205af9c1d28a239c0a105fcdc60ba800a70a
|
| a0cd2d701f29e0bc6275f1b13c0edfd4ec391879 |
|
01-Jun-2014 |
buzbee <buzbee@google.com> |
Quick compiler: reference cleanup
For 32-bit targets, object references are 32 bits wide both in
Dalvik virtual registers and in core physical registers. Because of
this, object references and non-floating point values were both
handled as if they had the same register class (kCoreReg).
However, for 64-bit systems, references are 32 bits in Dalvik vregs, but
64 bits in physical registers. Although the same underlying physical
core registers will still be used for object reference and non-float
values, different register class views will be used to represent them.
For example, an object reference in arm64 might be held in x3 at some
point, while the same underlying physical register, w3, would be used
to hold a 32-bit int.
This CL breaks apart the handling of object reference and non-float values
to allow the proper register class (or register view) to be used. A
new register class, kRefReg, is introduced which will map to a 32-bit
core register on 32-bit targets, and 64-bit core registers on 64-bit
targets. From this point on, object references should be allocated
registers in the kRefReg class rather than kCoreReg.
Change-Id: I6166827daa8a0ea3af326940d56a6a14874f5810
|
| ffddfdf6fec0b9d98a692e27242eecb15af5ead2 |
|
03-Jun-2014 |
Tim Murray <timmurray@google.com> |
DO NOT MERGE
Merge ART from AOSP to lmp-preview-dev.
Change-Id: I0f578733a4b8756fd780d4a052ad69b746f687a9
|
| 0955f7e470fb733aef07096536e9fba7c99250aa |
|
23-May-2014 |
Matteo Franchin <matteo.franchin@arm.com> |
AArch64: fixing some assertions.
Fixing some assertions while attempting to get libartd.so to work.
Fixing also the shift logic in LoadBaseIndexed() and StoreBaseIndexed().
This commit only fixes a part of the assertion issues.
Change-Id: I473194d4260dd59a8ee6d73114429728c977ee0e
|
| 85089dd28a39dd20f42ac258398b2a08668f9ef1 |
|
26-May-2014 |
buzbee <buzbee@google.com> |
Quick compiler: generalize NarrowRegLoc()
Some of the RegStorage utilites (DoubleToLowSingle(),
DoubleToHighSingle(), etc.) worked only for targets which
which treat double precision registers as a pair of aliased
single precision registers.
This CL elminates those utilities, and replaces them with
a new RegisterInfo utility that will search an aliased register
set and return the member matching the required storage
configuration (if it exists).
Change-Id: Iff5de10f467d20a56e1a89df9fbf30d1cf63c240
|
| 642fe34958ba7fafa81341823241616edde0380c |
|
24-May-2014 |
buzbee <buzbee@google.com> |
Quick compiler: fix register clobbering.
Ensure all aliased children of a register set are clobbered
when any member is clobbered. Additionally, use a clobbering
mask to avoid clobbering non-overlapping siblings.
Change-Id: Ic0d88a30f3e5b7a359396f6541d602739fa3124a
|
| ed65c5e982705defdb597d94d1aa3f2997239c9b |
|
22-May-2014 |
Serban Constantinescu <serban.constantinescu@arm.com> |
AArch64: Enable LONG_* and INT_* opcodes.
This patch fixes some of the issues with LONG and INT opcodes. The patch
has been tested and passes all the dalvik tests except for 018 and 107.
Change-Id: Idd1923ed935ee8236ab0c7e5fa969eaefeea8708
Signed-off-by: Serban Constantinescu <serban.constantinescu@arm.com>
|
| a51a0b0300268b605e3ad71b0e87ff394032c5e7 |
|
21-May-2014 |
Vladimir Marko <vmarko@google.com> |
Method inlining across dex files in boot image.
Fix LoadCodeAddress() and LoadMethodAddress() to use the dex
file in addition to the method index to uniquely identify
the literal. With that fix in place, when we have both the
direct code and the direct method, we can safely pass the
actual target method id instead of the method id from the
same dex file in the method lowering info. This was already
done for calls from apps into boot image (and thus there was
a bug with a tiny risk of the wrong literal being used) and
now we also do that for calls within the boot image. The
latter allows the inlining pass to inline many more methods
than before in the boot image.
Bug: 15021903
Change-Id: Ic765ce9809b43ef07e7db32b8e3fbc9acb09147f
|
| b01bf15d18f9b08d77e7a3c6e2897af0e02bf8ca |
|
14-May-2014 |
buzbee <buzbee@google.com> |
64-bit temp register support.
Add a 64-bit temp register allocation path. The recent physical
register handling rework supports multiple views of the same
physical register (or, such as for Arm's float/double regs,
different parts of the same physical register).
This CL adds a 64-bit core register view for 64-bit targets. In
short, each core register will have a 64-bit name, and a 32-bit
name. The different views will be kept in separate register pools,
but aliasing will be tracked. The core temp register allocation
routines will be largely identical - except for 32-bit targets,
which will continue to use pairs of 32-bit core registers for holding
long values.
Change-Id: I8f118e845eac7903ad8b6dcec1952f185023c053
|
| e87f9b5185379c8cf8392d65a63e7bf7e51b97e7 |
|
30-Apr-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Allow X86 QBE to be extended
Enhancements and updates to allow X86Mir2LIR Backend to be subclassed
for experimentation. Add virtual in a whole bunch of places, and make
some other changes to get this to work.
Change-Id: I0980a19bc5d5725f91660f98c95f1f51c17ee9b6
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 082833c8d577db0b2bebc100602f31e4e971613e |
|
18-May-2014 |
buzbee <buzbee@google.com> |
Quick compiler, out of registers fix
It turns out that the register pool sanity checker was not
working as expected, leaving some inconsistencies unreported.
This could result in "out of registers" failures, as well
as other more subtle problems.
This CL fixes the sanity checker, adds a lot more check and cleans
up the previously undetected episodes of insanity.
Cherry-pick of internal change 468162
Change-Id: Id2da97e99105a4c272c5fd256205a94b904ecea8
|
| 05d3aeb33683b16837741f9348d6fba9a8432068 |
|
18-May-2014 |
buzbee <buzbee@google.com> |
Quick compiler, out of registers fix
Fixes b/15024623
It turns out that the register pool sanity checker was not
working as expected, leaving some inconsistencies unreported.
This CL fixes the sanity checker, adds a lot more check and cleans
up the previously undetected episodes of insanity.
Change-Id: I4d67db864ca5926a1975db251e7e631b65a86275
|
| d65c51a556e6649db4e18bd083c8fec37607a442 |
|
29-Apr-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
ART: Add support for constant vector literals
Add in some vector instructions. Implement the ConstVector
instruction, which takes 4 words of data and loads it into
an XMM register.
Initially, only the ConstVector MIR opcode is implemented. Others will
be added after this one goes in.
Change-Id: I5c79bc8b7de9030ef1c213fc8b227debc47f6337
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| b14329f90f725af0f67c45dfcb94933a426d63ce |
|
15-May-2014 |
Andreas Gampe <agampe@google.com> |
ART: Fix MonitorExit code on ARM
We do not emit barriers on non-SMP systems. But on ARM, we have
places that need to conditionally execute, which is done through
an IT instruction. The guide of said instruction thus changes
between SMP and non-SMP systems.
To cleanly approach this, change the API so that GenMemBarrier
returns whether it generated an instruction. ARM will have to
query the result and update any dependent IT.
Throw a build system error if TARGET_CPU_SMP is not set.
Fix runtime/Android.mk to work with new multilib host.
Bug: 14989275
Change-Id: I9e611b770e8a1cd4ca19367d7dae0573ec08dc61
|
| c93ac8b73b5772e43b6dd1cc9e1deee79ca68849 |
|
13-May-2014 |
Vladimir Marko <vmarko@google.com> |
Fix special getter/setter to use RegClassForFieldLoadStore().
This ensures correct register class is used for volatile
load/store in these getters and setters.
Bug: 14112919
Change-Id: Ib7aa83d441fb007e97f9acc2a778bc20ffed837c
|
| 9ee801f5308aa3c62ae3bedae2658612762ffb91 |
|
12-May-2014 |
Dmitry Petrochenko <dmitry.petrochenko@intel.com> |
Add x86_64 code generation support
Utilizes r0..r7 in register allocator, implements spill/unsill
core regs as well as operations with stack pointer.
Change-Id: I973d5a1acb9aa735f6832df3d440185d9e896c67
Signed-off-by: Dmitry Petrochenko <dmitry.petrochenko@intel.com>
|
| 2f244e9faccfcca68af3c5484c397a01a1c3a342 |
|
08-May-2014 |
Andreas Gampe <agampe@google.com> |
ART: Add more ThreadOffset in Mir2Lir and backends
This duplicates all methods with ThreadOffset parameters, so that
both ThreadOffset<4> and ThreadOffset<8> can be handled. Dynamic
checks against the compilation unit's instruction set determine
which pointer size to use and therefore which methods to call.
Methods with unsupported pointer sizes should fatally fail, as
this indicates an issue during method selection.
Change-Id: Ifdb445b3732d3dc5e6a220db57374a55e91e1bf6
|
| ba57451494946a128703e1cbd8bf5969ee8dc598 |
|
13-May-2014 |
buzbee <buzbee@google.com> |
Quick compiler: fix compile-time perf regression
The recent changes to the temp register liveness tracking
introduced a measureable compile-time performance regression.
This CL cleans it up.
Change-Id: Id698b93e957f0ecab7ddfab94727f85e49cf10cf
|
| 0dc242d6fc1254e6ca1c31e08e612bbf45644b17 |
|
12-May-2014 |
Vladimir Marko <vmarko@google.com> |
Avoid unnecessary copy/load in EvalLoc() and LoadValue().
EvalLoc()/EvalLocWide() are used to prepare a register where
a value is subsequently stored, so they shouldn't copy the
old value to the new register for register class mismatch.
The only exception where we actually need a copy is
LoadValue()/LoadValueWide(), so we inline the old code that
makes the copy there. We also avoid loading inexpensive
constants when the value is already in the register.
Change-Id: I07519e9d4d9b3f7272233d196435f3035e4a3ca9
|
| 30adc7383a74eb3cb6db3bf42cea3a5595055ce1 |
|
10-May-2014 |
buzbee <buzbee@google.com> |
Quick compiler: Fix liveness tracking
Rework temp register liveness tracking to play nicely with aliased
physical registers, and re-enable liveness tracking optimization.
Add a pair of x86 utility routines that act like UpdateLoc(),
but only show in-register live temps if they are of the expected
register class.
Change-Id: I92779e0da2554689103e7488025be281f1a58989
|
| 674744e635ddbdfb311fbd25b5a27356560d30c3 |
|
24-Apr-2014 |
Vladimir Marko <vmarko@google.com> |
Use atomic load/store for volatile IGET/IPUT/SGET/SPUT.
Bug: 14112919
Change-Id: I79316f438dd3adea9b2653ffc968af83671ad282
|
| e45fb9e7976c8462b94a58ad60b006b0eacec49f |
|
06-May-2014 |
Matteo Franchin <matteo.franchin@arm.com> |
AArch64: Change arm64 backend to produce A64 code.
The arm backend clone is changed to produce A64 code. At the moment
this backend can only compile simple methods (both leaf and non-leaf).
Most of the work on the assembler (assembler_arm64.cc) has been done.
Some work on the LIR generation layer (functions such as OpRegRegImm
& friends) is still necessary. The register allocator still needs to
be adapted to the A64 instruction set (it is mostly unchanged from
the arm backend). Offsets for helpers in gen_invoke.cc still need to
be changed to work on 64-bit.
Change-Id: I388f99eeb832857981c7d9d5cb5b71af64a4b921
|
| 3bf7c60a86d49bf8c05c5d2ac5ca8e9f80bd9824 |
|
07-May-2014 |
Vladimir Marko <vmarko@google.com> |
Cleanup ARM load/store wide and remove unused param s_reg.
Use a single LDRD/VLDR instruction for wide load/store on
ARM, adjust the base pointer if needed. Remove unused
parameter s_reg from LoadBaseDisp(), LoadBaseIndexedDisp()
and StoreBaseIndexedDisp() on all architectures.
Change-Id: I25a9a42d523a68addbc11abe44ddc55a4401df98
|
| 455759b5702b9435b91d1b4dada22c4cce7cae3c |
|
06-May-2014 |
Vladimir Marko <vmarko@google.com> |
Remove LoadBaseDispWide and StoreBaseDispWide.
Just pass k64 or kDouble to non-wide versions.
Change-Id: I000619c3b78d3a71db42edc747c8a0ba1ee229be
|
| 091cc408e9dc87e60fb64c61e186bea568fc3d3a |
|
31-Mar-2014 |
buzbee <buzbee@google.com> |
Quick compiler: allocate doubles as doubles
Significant refactoring of register handling to unify usage across
all targets & 32/64 backends.
Reworked RegStorage encoding to allow expanded use of
x86 xmm registers; removed vector registers as a separate
register type. Reworked RegisterInfo to describe aliased
physical registers. Eliminated quite a bit of target-specific code
and generalized common code.
Use of RegStorage instead of int for registers now propagated down
to the NewLIRx() level. In future CLs, the NewLIRx() routines will
be replaced with versions that are explicit about what kind of
operand they expect (RegStorage, displacement, etc.). The goal
is to eventually use RegStorage all the way to the assembly phase.
TBD: MIPS needs verification.
TBD: Re-enable liveness tracking.
Change-Id: I388c006d5fa9b3ea72db4e37a19ce257f2a15964
|
| 6ffcfa04ebb2660e238742a6000f5ccebdd5df15 |
|
25-Apr-2014 |
Mingyao Yang <mingyao@google.com> |
Rewrite suspend test check with LIRSlowPath.
Change-Id: I2dc17d079655586bfc588349c7a04afc2c6879af
|
| 7a11ab09f93f54b1c07c0bf38dd65ed322e86bc6 |
|
29-Apr-2014 |
buzbee <buzbee@google.com> |
Quick compiler: debugging assists
A few minor assists to ease A/B debugging in the Quick
compiler:
1. To save time, the assemblers for some targets only
update the object code offsets on instructions involved with
pc-relative fixups. We add code to fix up all offsets when
doing a verbose codegen listing.
2. Temp registers are normally allocated in a round-robin
fashion. When disabling liveness tracking, we now reset the
round-robin pool to 0 on each instruction boundary. This makes
it easier to spot real codegen differences.
3. Self-register copies were previously emitted, but
marked as nops. Minor change to avoid generating them in the
first place and reduce clutter.
Change-Id: I7954bba3b9f16ee690d663be510eac7034c93723
|
| 695d13a82d6dd801aaa57a22a9d4b3f6db0d0fdb |
|
19-Apr-2014 |
buzbee <buzbee@google.com> |
Update load/store utilities for 64-bit backends
This CL replaces the typical use of LoadWord/StoreWord
utilities (which, in practice, were 32-bit load/store) in
favor of a new set that make the size explicit. We now have:
LoadWordDisp/StoreWordDisp:
32 or 64 depending on target. Load or store the natural
word size. Expect this to be used infrequently - generally
when we know we're dealing with a native pointer or flushed
register not holding a Dalvik value (Dalvik values will flush
to home location sizes based on Dalvik, rather than the target).
Load32Disp/Store32Disp:
Load or store 32 bits, regardless of target.
Load64Disp/Store64Disp:
Load or store 64 bits, regardless of target.
LoadRefDisp:
Load a 32-bit compressed reference, and expand it to the
natural word size in the target register.
StoreRefDisp:
Compress a reference held in a register of the natural word
size and store it as a 32-bit compressed reference.
Change-Id: I50fcbc8684476abd9527777ee7c152c61ba41c6f
|
| 3a74d15ccc9a902874473ac9632e568b19b91b1c |
|
22-Apr-2014 |
Mingyao Yang <mingyao@google.com> |
Delete throw launchpads.
Bug: 13170824
Change-Id: I9d5834f5a66f5eb00f2ac80774e8c27dea99949e
|
| 80365d9bb947edef0eae0bfe62b9f7a239416e6b |
|
18-Apr-2014 |
Mingyao Yang <mingyao@google.com> |
Revert "Revert "Use LIRSlowPath for throwing ArrayOutOfBoundsException.""
This adds back using LIRSlowPath for ArrayIndexOutOfBoundsException.
And fix the host test crash.
Change-Id: Idbb602f4bb2c5ce59233feb480a0ff1b216e4887
|
| 7fff544c38f0dec3a213236bb785c3ca13d21a0f |
|
18-Apr-2014 |
Brian Carlstrom <bdc@google.com> |
Revert "Use LIRSlowPath for throwing ArrayOutOfBoundsException."
This reverts commit 9d46314a309aff327f9913789b5f61200c162609.
|
| 9d46314a309aff327f9913789b5f61200c162609 |
|
18-Apr-2014 |
Mingyao Yang <mingyao@google.com> |
Use LIRSlowPath for throwing ArrayOutOfBoundsException.
Get rid of launchpads for throwing ArrayOutOfBoundsException
and use LIRSlowPath instead.
Bug: 13170824
Change-Id: I0e27f7a261a6a7fb5c0645e6113a957e098f699e
|
| e643a179cf5585ba6bafdd4fa51730d9f50c06f6 |
|
08-Apr-2014 |
Mingyao Yang <mingyao@google.com> |
Use LIRSlowPath for throwing NPE.
Get rid of launchpads for throwing NPE and use LIRSlowPath instead.
Also clean up some code of using LIRSlowPath for checking div
by zero.
Bug: 13170824
Change-Id: I0c20a49c39feff3eb1f147755e557d9bc0ff15bb
|
| d6ed642458c8820e1beca72f3d7b5f0be4a4b64b |
|
10-Apr-2014 |
Dave Allison <dallison@google.com> |
Revert "Revert "Revert "Use trampolines for calls to helpers"""
This reverts commit f9487c039efb4112616d438593a2ab02792e0304.
Change-Id: Id48a4aae4ecce73db468587967968a3f7618b700
|
| f9487c039efb4112616d438593a2ab02792e0304 |
|
09-Apr-2014 |
Dave Allison <dallison@google.com> |
Revert "Revert "Use trampolines for calls to helpers""
This reverts commit 081f73e888b3c246cf7635db37b7f1105cf1a2ff.
Change-Id: Ibd777f8ce73cf8ed6c4cb81d50bf6437ac28cb61
Conflicts:
compiler/dex/quick/mir_to_lir.h
|
| 4289456fa265b833434c2a8eee9e7a16da31c524 |
|
07-Apr-2014 |
Mingyao Yang <mingyao@google.com> |
Use LIRSlowPath for throwing div by zero exception.
Get rid of launchpads for throwing div by zero exception and
use LIRSlowPath instead. Add a CallRuntimeHelper that takes no
argument for the runtime function.
Bug: 13170824
Change-Id: I7e0563e736c6f92bd63e3fbdfe3a777ad333e338
|
| 081f73e888b3c246cf7635db37b7f1105cf1a2ff |
|
07-Apr-2014 |
Dave Allison <dallison@google.com> |
Revert "Use trampolines for calls to helpers"
This reverts commit 754ddad084ccb610d0cf486f6131bdc69bae5bc6.
Change-Id: Icd979adee1d8d781b40a5e75daf3719444cb72e8
|
| 754ddad084ccb610d0cf486f6131bdc69bae5bc6 |
|
19-Feb-2014 |
Dave Allison <dallison@google.com> |
Use trampolines for calls to helpers
This is an ARM specific optimization to the compiler
that uses trampoline islands to make calls to runtime
helper functions. The intention is to reduce the size
of the generated code (by 2 bytes per call) without
affecting performance.
By default this is on when generating an OAT file. It is
off when compiling to memory.
To switch this off in dex2oat, use the command line option:
--no-helper-trampolines
Enhances disassembler to print the trampoline entry on the
BL instruction like this:
0xb6a850c0: f7ffff9e bl -196 (0xb6a85000) ; pTestSuspend
Bug: 12607709
Change-Id: I9202bdb7cf21252ad807bd48701f1f6ce8e3d0fe
|
| 3da67a558f1fd3d8a157d8044d521753f3f99ac8 |
|
03-Apr-2014 |
Dave Allison <dallison@google.com> |
Add OpEndIT() for marking the end of OpIT blocks
In ARM we need to prevent code motion to the inside of an
IT block. This was done using a GenBarrier() to mark the end, but
it wasn't obvious that this is what was happening. This CL adds
an explicit OpEndIT() that takes the LIR of the OpIT for future
checks.
Bug: 13751744
Change-Id: If41d2adea1f43f11ebb3b72906bd308252ce3d01
|
| dd7624d2b9e599d57762d12031b10b89defc9807 |
|
15-Mar-2014 |
Ian Rogers <irogers@google.com> |
Allow mixing of thread offsets between 32 and 64bit architectures.
Begin a more full implementation x86-64 REX prefixes.
Doesn't implement 64bit thread offset support for the JNI compiler.
Change-Id: If9af2f08a1833c21ddb4b4077f9b03add1a05147
|
| f943914730db8ad2ff03d49a2cacd31885d08fd7 |
|
27-Mar-2014 |
Dave Allison <dallison@google.com> |
Implement implicit stack overflow checks
This also fixes some failing run tests due to missing
null pointer markers.
The implementation of the implicit stack overflow checks introduces
the ability to have a gap in the stack that is skipped during
stack walk backs. This gap is protected against read/write and
is used to trigger a SIGSEGV at function entry if the stack
will overflow.
Change-Id: I0c3e214c8b87dc250cf886472c6d327b5d58653e
|
| 306f017dd883c0bf806d239d97e0bca3194afbd7 |
|
07-Jan-2014 |
Vladimir Marko <vmarko@google.com> |
Faster AssembleLIR for ARM.
This also reduces sizeof(LIR) by 4 bytes (32-bit builds).
Change-Id: I0cb81f9bf098dfc50050d5bc705c171af26464ce
|
| e2143c0a4af68c08e811885eb2f3ea5bfdb21ab6 |
|
28-Mar-2014 |
Ian Rogers <irogers@google.com> |
Revert "Revert "Optimize easy multiply and easy div remainder.""
This reverts commit 3654a6f50a948ead89627f398aaf86a2c2db0088.
Remove the part of the change that confused !is_div with being multiply rather
than implying remainder.
Change-Id: I202610069c69351259a320e8852543cbed4c3b3e
|
| 9da5c1013215176f2a4dbe7a804be899e12d5f68 |
|
28-Mar-2014 |
buzbee <buzbee@google.com> |
Quick compiler, MIPS resource cleanup
MIPS architecture includes internal registers HI and LO.
Similar to condition codes in other architectures, these internal
resouces must be accounted for during instruction scheduling.
Previously, the Quick backend for MIPS dealt with them by defining
rHI and rLO pseudo registers - treating them as actual registers for
def/use masks. This CL changes the handling of these resources to
be in line with how condition codes are used elsewhere - leaving
register definitions to be used for registers.
Change-Id: Idcd77f3107b0c9b081ad05b1aab663fb9f41492d
|
| 3441512d61ac192c1bf0b9b1eb696d5a8a8d677e |
|
28-Mar-2014 |
Brian Carlstrom <bdc@google.com> |
Revert "Optimize easy multiply and easy div remainder."
This reverts commit 08df4b3da75366e5db37e696eaa7e855cba01deb.
(cherry picked from commit 3654a6f50a948ead89627f398aaf86a2c2db0088)
Change-Id: If8befd7c7135b9dfe3d3e9111768aba89aaa0863
|
| 3654a6f50a948ead89627f398aaf86a2c2db0088 |
|
28-Mar-2014 |
Brian Carlstrom <bdc@google.com> |
Revert "Optimize easy multiply and easy div remainder."
This reverts commit 08df4b3da75366e5db37e696eaa7e855cba01deb.
|
| 262b299abf658c16f61dad2240cfaf3deafe4423 |
|
27-Mar-2014 |
buzbee <buzbee@google.com> |
Fix x86 master build failure.
Replace bogus DCHECKs with logic matching pre-cleanup code.
Register pairs are considered temp, promoted, dirty or live if
either register of the pair meets criteria.
Change-Id: If2df891fdd1e3351d4cbe72aaf2a2ac5b34b2110
|
| 14a46d820b04b848063f7c32ecd2cf82dd90cb1d |
|
27-Mar-2014 |
buzbee <buzbee@google.com> |
Fix x86 master build failure.
Replace bogus DCHECKs with logic matching pre-cleanup code.
Register pairs are considered temp, promoted, dirty or live if
either register of the pair meets criteria.
Change-Id: If2df891fdd1e3351d4cbe72aaf2a2ac5b34b2110
|
| 08df4b3da75366e5db37e696eaa7e855cba01deb |
|
25-Mar-2014 |
Zheng Xu <zheng.xu@arm.com> |
Optimize easy multiply and easy div remainder.
Update OpRegRegShift and OpRegRegRegShift to use RegStorage parameters.
Add special cases for *0 and *1. Add more easy multiply special cases for
Arm.
Reuse easy multiply in SmallLiteralDivRem() to support remainder cases.
Change-Id: Icd76a993d3ac8d4988e9653c19eab4efca14fad0
|
| 2700f7e1edbcd2518f4978e4cd0e05a4149f91b6 |
|
07-Mar-2014 |
buzbee <buzbee@google.com> |
Continuing register cleanup
Ready for review.
Continue the process of using RegStorage rather than
ints to hold register value in the top layers of codegen.
Given the huge number of changes in this CL, I've attempted
to minimize the number of actual logic changes. With this
CL, the use of ints for registers has largely been eliminated
except in the lowest utility levels. "Wide" utility routines
have been updated to take a single RegStorage rather than
a pair of ints representing low and high registers.
Upcoming CLs will be smaller and more targeted. My expectations:
o Allocate float double registers as a single double rather than
a pair of float single registers.
o Refactor to push code which assumes long and double Dalvik
values are held in a pair of register to the target dependent
layer.
o Clean-up of the xxx_mir.h files to reduce the amount of #defines
for registers. May also do a register renumbering to bring all
of our targets' register naming more consistent. Possibly
introduce a target-independent float/non-float test at the
RegStorage level.
Change-Id: I646de7392bdec94595dd2c6f76e0f1c4331096ff
|
| 99ad7230ccaace93bf323dea9790f35fe991a4a2 |
|
26-Feb-2014 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
Relaxed memory barriers for x86
X86 provides stronger memory guarantees and thus the memory barriers can be
optimized. This patch ensures that all memory barriers for x86 are treated
as scheduling barriers. And in cases where a barrier is needed (StoreLoad case),
an mfence is used.
Change-Id: I13d02bf3f152083ba9f358052aedb583b0d48640
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| 60d7a65f7fb60f502160a2e479e86014c7787553 |
|
14-Mar-2014 |
Brian Carlstrom <bdc@google.com> |
Fix stack overflow for mutual recursion.
There was an error where we would have a pc that was in the method
which generated the stack overflow. This didn't work however
because the stack overflow check was before we stored the method in
the stack. The result was that the stack overflow handler had a PC
which wasnt necessarily in the method at the top of the stack. This
is now fixed by always restoring the link register before branching
to the throw entrypoint.
Slight code size regression on ARM/Mips (unmeasured). Regression on ARM
is 4 bytes of code per stack overflow check. Some of this regression is
mitigated by having one less GC safepoint.
Also adds test case for StackOverflowError issue (from bdc).
Tests passing: ARM, X86, Mips
Phone booting: ARM
Bug: https://code.google.com/p/android/issues/detail?id=66411
Bug: 12967914
Change-Id: I96fe667799458b58d1f86671e051968f7be78d5d
(cherry-picked from c0f96d03a1855fda7d94332331b94860404874dd)
|
| c0f96d03a1855fda7d94332331b94860404874dd |
|
14-Mar-2014 |
Brian Carlstrom <bdc@google.com> |
Fix stack overflow for mutual recursion.
There was an error where we would have a pc that was in the method
which generated the stack overflow. This didn't work however
because the stack overflow check was before we stored the method in
the stack. The result was that the stack overflow handler had a PC
which wasnt necessarily in the method at the top of the stack. This
is now fixed by always restoring the link register before branching
to the throw entrypoint.
Slight code size regression on ARM/Mips (unmeasured). Regression on ARM
is 4 bytes of code per stack overflow check. Some of this regression is
mitigated by having one less GC safepoint.
Also adds test case for StackOverflowError issue (from bdc).
Tests passing: ARM, X86, Mips
Phone booting: ARM
Bug: https://code.google.com/p/android/issues/detail?id=66411
Bug: 12967914
Change-Id: I96fe667799458b58d1f86671e051968f7be78d5d
|
| e90501da0222717d75c126ebf89569db3976927e |
|
12-Mar-2014 |
Serguei Katkov <serguei.i.katkov@intel.com> |
Add dependency for operations with x86 FPU stack
Load Hoisting optimization can re-order operations with
FPU stack due to no dependency set.
Patch adds resource dependency between these operations.
Change-Id: Iccce98c8f3c565903667c03803884d9de1281ea8
Signed-off-by: Serguei Katkov <serguei.i.katkov@intel.com>
|
| b373e091eac39b1a79c11f2dcbd610af01e9e8a9 |
|
21-Feb-2014 |
Dave Allison <dallison@google.com> |
Implicit null/suspend checks (oat version bump)
This adds the ability to use SEGV signals
to throw NullPointerException exceptions from Java code rather
than having the compiler generate explicit comparisons and
branches. It does this by using sigaction to trap SIGSEGV and when triggered
makes sure it's in compiled code and if so, sets the return
address to the entry point to throw the exception.
It also uses this signal mechanism to determine whether to check
for thread suspension. Instead of the compiler generating calls
to a function to check for threads being suspended, the compiler
will now load indirect via an address in the TLS area. To trigger
a suspend, the contents of this address are changed from something
valid to 0. A SIGSEGV will occur and the handler will check
for a valid instruction pattern before invoking the thread
suspension check code.
If a user program taps SIGSEGV it will prevent our signal handler
working. This will cause a failure in the runtime.
There are two signal handlers at present. You can control them
individually using the flags -implicit-checks: on the runtime
command line. This takes a string parameter, a comma
separated set of strings. Each can be one of:
none switch off
null null pointer checks
suspend suspend checks
all all checks
So to switch only suspend checks on, pass:
-implicit-checks:suspend
There is also -explicit-checks to provide the reverse once
we change the default.
For dalvikvm, pass --runtime-arg -implicit-checks:foo,bar
The default is -implicit-checks:none
There is also a property 'dalvik.vm.implicit_checks' whose value is the same
string as the command option. The default is 'none'. For example to switch on
null checks using the option:
setprop dalvik.vm.implicit_checks null
It only works for ARM right now.
Bumps OAT version number due to change to Thread offsets.
Bug: 13121132
Change-Id: If743849138162f3c7c44a523247e413785677370
|
| 3bc8615332b7848dec8c2297a40f7e4d176c0efb |
|
13-Mar-2014 |
Vladimir Marko <vmarko@google.com> |
Use LIRSlowPath for intrinsics, improve String.indexOf().
Rewrite intrinsic launchpads to use the LIRSlowPath.
Improve String.indexOf for constant chars by avoiding
the check for code points over 0xFFFF.
Change-Id: I7fd5583214c5b4ab9c38ee36c5d6f003dd6345a8
|
| 49161cef10a308aedada18e9aa742498d6e6c8c7 |
|
12-Mar-2014 |
Jeff Hao <jeffhao@google.com> |
Allow patching between dex files in the boot classpath.
Change-Id: I53f219a5382d0fcd580e96e50025fdad4fc399df
|
| 83cc7ae96d4176533dd0391a1591d321b0a87f4f |
|
12-Feb-2014 |
Vladimir Marko <vmarko@google.com> |
Create a scoped arena allocator and use that for LVN.
This saves more than 0.5s of boot.oat compilation time
on Nexus 5.
TODO: Move other stuff to the scoped allocator. This CL
alone increases the peak memory allocation. By reusing
the memory for other parts of the compilation we should
reduce this overhead.
Change-Id: Ifbc00aab4f3afd0000da818dfe68b96713824a08
|
| a1a7074eb8256d101f7b5d256cda26d7de6ce6ce |
|
03-Mar-2014 |
Vladimir Marko <vmarko@google.com> |
Rewrite kMirOpSelect for all IF_ccZ opcodes.
Also improve special cases for ARM and add tests.
Change-Id: I06f575b9c7b547dbc431dbfadf2b927151fe16b9
|
| 00e1ec6581b5b7b46ca4c314c2854e9caa647dd2 |
|
28-Feb-2014 |
Bill Buzbee <buzbee@android.com> |
Revert "Revert "Rework Quick compiler's register handling""
This reverts commit 86ec520fc8b696ed6f164d7b756009ecd6e4aace.
Ready. Fixed the original type, plus some mechanical changes
for rebasing.
Still needs additional testing, but the problem with the original
CL appears to have been a typo in the definition of the x86
double return template RegLocation.
Change-Id: I828c721f91d9b2546ef008c6ea81f40756305891
|
| be0e546730e532ef0987cd4bde2c6f5a1b14dd2a |
|
26-Feb-2014 |
Vladimir Marko <vmarko@google.com> |
Cache field lowering info in mir_graph.
Change-Id: I9f9d76e3ae6c31e88bdf3f59820d31a625da020f
|
| 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>
|
| a1ce1fef2d49d1d537776a5308ace7102a815fe5 |
|
25-Feb-2014 |
Brian Carlstrom <bdc@google.com> |
Split up CommonTest into CommonRuntimeTest and CommonCompilerTest
Change-Id: I8dcf6b29a5aecd445f1a3ddb06386cf81dbc9c70
|
| 86ec520fc8b696ed6f164d7b756009ecd6e4aace |
|
26-Feb-2014 |
Bill Buzbee <buzbee@android.com> |
Revert "Rework Quick compiler's register handling"
This reverts commit 2c1ed456dcdb027d097825dd98dbe48c71599b6c.
Change-Id: If88d69ba88e0af0b407ff2240566d7e4545d8a99
|
| 2c1ed456dcdb027d097825dd98dbe48c71599b6c |
|
20-Feb-2014 |
buzbee <buzbee@google.com> |
Rework Quick compiler's register handling
For historical reasons, the Quick backend found it convenient
to consider all 64-bit Dalvik values held in registers
to be contained in a pair of 32-bit registers. Though this
worked well for ARM (with double-precision registers also
treated as a pair of 32-bit single-precision registers) it doesn't
play well with other targets. And, it is somewhat problematic
for 64-bit architectures.
This is the first of several CLs that will rework the way the
Quick backend deals with physical registers. The goal is to
eliminate the "64-bit value backed with 32-bit register pair"
requirement from the target-indendent portions of the backend
and support 64-bit registers throughout.
The key RegLocation struct, which describes the location of
Dalvik virtual register & register pairs, previously contained
fields for high and low physical registers. The low_reg and
high_reg fields are being replaced with a new type: RegStorage.
There will be a single instance of RegStorage for each RegLocation.
Note that RegStorage does not increase the space used. It is
16 bits wide, the same as the sum of the 8-bit low_reg and
high_reg fields.
At a target-independent level, it will describe whether the physical
register storage associated with the Dalvik value is a single 32
bit, single 64 bit, pair of 32 bit or vector. The actual register
number encoding is left to the target-dependent code layer.
Because physical register handling is pervasive throughout the
backend, this restructuring necessarily involves large CLs with
lots of changes. I'm going to roll these out in stages, and
attempt to segregate the CLs with largely mechanical changes from
those which restructure or rework the logic.
This CL is of the mechanical change variety - it replaces low_reg
and high_reg from RegLocation and introduces RegStorage. It also
includes a lot of new code (such as many calls to GetReg())
that should go away in upcoming CLs.
The tentative plan for the subsequent CLs is:
o Rework standard register utilities such as AllocReg() and
FreeReg() to use RegStorage instead of ints.
o Rework the target-independent GenXXX, OpXXX, LoadValue,
StoreValue, etc. routines to take RegStorage rather than
int register encodings.
o Take advantage of the vector representation and eliminate
the current vector field in RegLocation.
o Replace the "wide" variants of codegen utilities that take
low_reg/high_reg pairs with versions that use RegStorage.
o Add 64-bit register target independent codegen utilities
where possible, and where not virtualize with 32-bit general
register and 64-bit general register variants in the target
dependent layer.
o Expand/rework the LIR def/use flags to allow for more registers
(currently, we lose out on 16 MIPS floating point regs as
well as ARM's D16..D31 for lack of space in the masks).
o [Possibly] move the float/non-float determination of a register
from the target-dependent encoding to RegStorage. In other
words, replace IsFpReg(register_encoding_bits).
At the end of the day, all code in the target independent layer
should be using RegStorage, as should much of the target dependent
layer. Ideally, we won't be using the physical register number
encoding extracted from RegStorage (i.e. GetReg()) until the
NewLIRx() layer.
Change-Id: Idc5c741478f720bdd1d7123b94e4288be5ce52cb
|
| 9c86a0279aaf953377aa9e2277592e68bf814989 |
|
21-Feb-2014 |
Ian Rogers <irogers@google.com> |
Revert "Annotate used fields."
This reverts commit 7f6cf56942c8469958b273ea968db253051c5b05.
Change-Id: Ic389a194c3404ecb5bb563a405bf4a0d6336ea0d
|
| 4028a6c83a339036864999fdfd2855b012a9f1a7 |
|
20-Feb-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Inline x86 String.indexOf
Take advantage of the presence of a constant search char or start index
to tune the generated code.
Change-Id: I0adcf184fb91b899a95aa4d8ef044a14deb51d88
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 7f6cf56942c8469958b273ea968db253051c5b05 |
|
29-Jan-2014 |
Vladimir Marko <vmarko@google.com> |
Annotate used fields.
Annotate all fields used by a method early during the
compilation, check acces rights and record field offset,
volatility, etc. Use these annotations when generating code
for IGET/IPUT/SGET/SPUT instructions.
Change-Id: I4bbf5cca4fecf53c9bf9c93ac1793e2f40c16b5f
|
| 818f2107e6d2d9e80faac8ae8c92faffa83cbd11 |
|
18-Feb-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Re-apply: Initial check-in of an optimizing compiler.
The classes and the names are very much inspired by V8/Dart.
It currently only supports the RETURN_VOID dex instruction,
and there is a pretty printer to check if the building of the
graph is correct.
Change-Id: I28e125dfee86ae6ec9b3fec6aa1859523b92a893
|
| 1af0c0b88a956813eb0ad282664cedc391e2938f |
|
19-Feb-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Revert "Initial check-in of an optimizing compiler."
g++ warnings turned into errors.
This reverts commit 68a5fefa90f03fdf5a238ac85c9439c6b03eae96.
Change-Id: I09bb95d9cc13764ca8a266c41af04801a34b9fd0
|
| 68a5fefa90f03fdf5a238ac85c9439c6b03eae96 |
|
18-Feb-2014 |
Nicolas Geoffray <ngeoffray@google.com> |
Initial check-in of an optimizing compiler.
The classes and the names are very much inspired by V8/Dart.
It currently only supports the RETURN_VOID dex instruction,
and there is a pretty printer to check if the building of the
graph is correct.
Change-Id: Id5ef1b317ab997010d4e3888e456c26bef1ab9c0
|
| 3bc01748ef1c3e43361bdf520947a9d656658bf8 |
|
06-Feb-2014 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
GenSpecialCase support for x86
Moved GenSpecialCase from being ARM specific to common code to allow
it to be used by x86 quick as well.
Change-Id: I728733e8f4c4da99af6091ef77e5c76ae0fee850
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| 614c2b4e219631e8c190fd9fd5d4d9cd343434e1 |
|
29-Jan-2014 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
Support to generate inline long to FP bytecodes for x86
long-to-float and long-to-double are now generated inline instead of calling
a helper routine. The conversion is done by using x87.
Change-Id: I196e526afec1be212898baceca8527549c3655b6
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| 6607d97166984ce578817269f9775c15b9044190 |
|
10-Feb-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Tweak Mir2Lir::GenInstanceofCallingHelper for X86
Make this virtual, and split out the X86 logic. Take advantage of SETcc
instruction for X86.
I don't think I can do much more due to need to preserve arguments for
the calls.
Change-Id: I10e3eaa61b61ceac384267e3078bb6f75c37cee4
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 55d0eac918321e0525f6e6491f36a80977e0d416 |
|
06-Feb-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Support Direct Method/Type access for X86
Thumb generates code to optimize calls to methods within core.oat.
Implement this for X86 as well, but take advantage of mov with 32 bit
immediate and call relative with 32 bit immediate.
Fix some incorrect return locations for long inlines.
Change-Id: I1907bdfc7574f3d0aa76c7fad13dc537acdf1ed3
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| dbb17e378b538133750e56375bbdbb217db7b248 |
|
07-Feb-2014 |
Yixin Shou <yixin.shou@intel.com> |
Added inlined abs method with float and double type
This patch added the implementation for inlining java.lang.Math.abs()
method with float and double type.
Change-Id: Ic99471b4ab4176e4a0153bef383bb49944fb636f
Signed-off-by: Yixin Shou <yixin.shou@intel.com>
|
| 2c498d1f28e62e81fbdb477ff93ca7454e7493d7 |
|
30-Jan-2014 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
Specializing x86 range argument copying
The ARM implementation of range argument copying was specialized in some cases.
For all other architectures, it would fall back to generating memcpy. This patch
updates the x86 implementation so it does not call memcpy and instead generates
loads and stores, favoring movement of 128-bit chunks.
Change-Id: Ic891e5609a4b0e81a47c29cc5a9b301bd10a1933
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| bcec6fba95ee7974d3f7b81c3c02e7eb3ca3df00 |
|
17-Jan-2014 |
Dave Allison <dallison@google.com> |
Make slow paths easier to write
This adds a class LIRSlowPath that allows for deferred compilation
of slow paths. Using this object you can add code that will be
invoked out of line using a forward branch. The intention is to
move the slow paths out of the main flow and avoid branch-over
constructs that will almost always trigger. The forward branch
to the slow path code will be predicted false and this will
be correct most of the time. The slow path code returns to the
instruction after the original branch using an unconditional branch.
This is used in the following opcodes: sput, sget, const-string,
check-cast, const-class.
Others will follow.
Bug: 10864890
Change-Id: I17130c5dc20d369bc6bbf50b8cf04343263e888e
|
| feb2b4e2d1c6538777bb80b60f3a247537b6221d |
|
28-Jan-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Redo x86 int arithmetic
Make Mir2Lir::GenArithOpInt virtual, and implement an x86 version of it
to allow use of memory operands and knowledge of the fact that x86 has
(mostly) two operand instructions. Remove x86 specific code from the
generic version.
Add StoreFinalValue (matches StoreFinalValueWide) to handle the non-wide
cases. Add some x86 helper routines to simplify generation.
Change-Id: I6c13689c6da981f2570ab5af7a97f9816108b7ae
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| da7a69b3fa7bb22d087567364b7eb5a75824efd8 |
|
09-Jan-2014 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
Enable compiler temporaries
Compiler temporaries are a facility for having virtual register sized space
for dealing with intermediate values during MIR transformations. They receive
explicit space in managed frames so they can have a home location in case they
need to be spilled. The facility also supports "special" temporaries which
have specific semantic purpose and their location in frame must be tracked.
The compiler temporaries are treated in the same way as virtual registers
so that the MIR level transformations do not need to have special logic. However,
generated code needs to know stack layout so that it can distinguish between
home locations.
MIRGraph has received an interface for dealing with compiler temporaries. This
interface allows allocation of wide and non-wide virtual register temporaries.
The information about how temporaries are kept on stack has been moved to
stack.h. This is was necessary because stack layout is dependent on where the
temporaries are placed.
Change-Id: Iba5cf095b32feb00d3f648db112a00209c8e5f55
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| df8ee2ea9908db3dde463fed68391b0040517653 |
|
28-Jan-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
x86 updates GenInlinedUnsafePut/GenInstanceofFinal
Allow x86 to inline GenInlinedUnsafePut by freeing up a temporary
register early. Make an x86 specific version of GenInstanceofFinal that
uses compare to memory and a setCC instruction.
Change-Id: I67788d7ae83776b0b9069fe4b379452190774992
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 766e9295d2c34cd1846d81610c9045b5d5093ddd |
|
27-Jan-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Improve GenConstString, GenS{get,put} for x86
Rewrite GenConstString for x86 to skip calling ResolveString when the
string is already resolved. Also try to avoid a register copy if the
Method* is in a promoted register.
Implement the TODO for GenS{get,put} to use compare to memory for x86 by
adding a new codegen function to compare directly to memory. Implement
a default implementation that uses a temporary register for RISC
architectures.
Change-Id: Ie163cca3d3d841aa10c50dc6592ec30af7a7cbc9
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| bb8f0ab736b61db8f543e433859272e83f96ee9b |
|
28-Jan-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Embed array class pointers at array allocation sites.
Following https://android-review.googlesource.com/#/c/79302, embed
array class pointers at array allocation sites in the compiled code.
Change-Id: I67a1292466dfbb7f48e746e5060e992dd93525c5
|
| e27b3bf2c1044bfbfbe874affd3758a73009c6c6 |
|
23-Jan-2014 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
Support GenSelect for x86
kMirOpSelect is an extended MIR that has been generated in order
to remove trivial diamond shapes where the conditional is an
if-eqz or if-nez and on each of the paths there is a move or
const bytecode with same destination register.
This patch enables x86 to generate code for this extended MIR.
A) Handling the constant specialization of kMirOpSelect:
1) When the true case is zero and result_reg is not same as src_reg:
xor result_reg, result_reg
cmp $0, src_reg
mov t1, $false_case
cmovnz result_reg, t1
2) When the false case is zero and result_reg is not same as src_reg:
xor result_reg, result_reg
cmp $0, src_reg
mov t1, $true_case
cmovz result_reg, t1
3) All other cases (we do compare first to set eflags):
cmp $0, src_reg
mov result_reg, $true_case
mov t1, $false_case
cmovnz result_reg, t1
B) Handling the move specialization of kMirOpSelect:
1) When true case is already in place:
cmp $0, src_reg
cmovnz result_reg, false_reg
2) When false case is already in place:
cmp $0, src_reg
cmovz result_reg, true_reg
3) When neither cases are in place:
cmp $0, src_reg
mov result_reg, true_reg
cmovnz result_reg, false_reg
Change-Id: Ic7c50823208fe82019916476a0a77c6a271679fe
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| 4708dcd68eebf1173aef1097dad8ab13466059aa |
|
22-Jan-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Improve x86 long multiply and shifts
Generate inline code for long shifts by constants and do long
multiplication inline. Convert multiplication by a constant to a
shift when we can. Fix some x86 assembler problems and add the new
instructions that were needed (64 bit shifts).
Change-Id: I6237a31c36159096e399d40d01eb6bfa22ac2772
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| 2bf31e67694da24a19fc1f328285cebb1a4b9964 |
|
23-Jan-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Improve x86 long divide
Implement inline division for literal and variable divisors. Use the
general case for dividing by a literal by using a double length multiply
by the appropriate constant with fixups. This is the Hacker's Delight
algorithm.
Change-Id: I563c250f99d89fca5ff8bcbf13de74de13815cfe
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| be1ca55db3362f5b100c4c65da5342fd299520bb |
|
15-Jan-2014 |
Hiroshi Yamauchi <yamauchi@google.com> |
Use direct class pointers at allocation sites in the compiled code.
- Rather than looking up a class from its type ID (and checking if
it's resolved/initialized, resolving/initializing if not), use
direct class pointers, if possible (boot-code-to-boot-class pointers
and app-code-to-boot-class pointers.)
- This results in a 1-2% speedup in Ritz MemAllocTest on Nexus 4.
- Embedding the object size (along with class pointers) caused a 1-2%
slowdown in MemAllocTest and isn't implemented in this change.
- TODO: do the same for array allocations.
- TODO: when/if an application gets its own image, implement
app-code-to-app-class pointers.
- Fix a -XX:gc bug.
cf. https://android-review.googlesource.com/79460/
- Add /tmp/android-data/dalvik-cache to the list of locations to
remove oat files in clean-oat-host.
cf. https://android-review.googlesource.com/79550
- Add back a dropped UNLIKELY in FindMethodFromCode().
cf. https://android-review.googlesource.com/74205
Bug: 9986565
Change-Id: I590b96bd21f7a7472f88e36752e675547559a5b1
|
| e02d48fb24747f90fd893e1c3572bb3c500afced |
|
15-Jan-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Optimize x86 long arithmetic
Be smarter about taking advantage of a constant operand for x86 long
add/sub/and/or/xor. Using instructions with immediates and generating
results directly into memory reduces the number of temporary registers
and avoids hardcoded register usage.
Also rewrite the existing non-const x86 arithmetic to avoid fixed
register use, and use the fact that x86 instructions are two operand.
Pass the opcode to the XXXLong() routines to easily detect two operand
DEX opcodes.
Add a new StoreFinalValueWide() routine, which is similar to StoreValueWide,
but doesn't do an EvalLoc to allocate registers. The src operand must
already be in registers, and it just updates the dest location, and
calls the right live/dirty routines to get the src into the dest
properly.
Change-Id: Iefc16e7bc2236a73dc780d3d5137ae8343171f62
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| d61ba4ba6fcde666adb5d5c81b1c32f0534fb2c8 |
|
13-Jan-2014 |
Bill Buzbee <buzbee@android.com> |
Revert "Revert "Better support for x86 XMM registers""
This reverts commit 8ff67e3338952c70ccf3b609559bf8cc0f379cfd.
Fix applied to loc.fp usage.
Change-Id: I1eb3005392544fcf30c595923ed25bcee2dc4859
|
| 8ff67e3338952c70ccf3b609559bf8cc0f379cfd |
|
11-Jan-2014 |
Bill Buzbee <buzbee@android.com> |
Revert "Better support for x86 XMM registers"
The invalid usage of loc.fp must be corrected before this change can be submitted.
This reverts commit 766a5e5940b469ab40e52770862c81cfec1d835b.
Change-Id: I1173a9bf829da89cccd9c2898f5e11164987a22b
|
| 766a5e5940b469ab40e52770862c81cfec1d835b |
|
10-Jan-2014 |
Mark Mendell <mark.p.mendell@intel.com> |
Better support for x86 XMM registers
Currently, ART Quick mode assumes that a double FP register is composed
of two single consecutive FP registers. This is true for ARM and MIPS,
but not x86. This means that only half of the 8 XMM registers are
available for use by x86 doubles.
This patch breaks the assumption that a wide FP RegisterLocation must be
a paired set of FP registers. This is done by making some routines in
common code virtual and overriding them in the X86Mir2Lir class. For
these wide fp locations, the high register is set to the same value as
the low register, in order to minimize changes to common code. In a
couple of places, the common code checks for this case.
The changes are also supposed to allow the possibility of using the XMM
registers for vector operations,but that support is still WIP.
Change-Id: Ic6ef24ea764991c6f4d9fb88d483a619f5a468cb
Signed-off-by: Mark Mendell <mark.p.mendell@intel.com>
|
| bd288c2c1206bc99fafebfb9120a83f13cf9723b |
|
21-Dec-2013 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
Add conditional move support to x86 and allow GenMinMax to use it
X86 supports conditional moves which is useful for reducing branchiness.
This patch adds support to the x86 backend to generate conditional reg
to reg operations. Both encoder and decoder support was added for cmov.
The x86 version of GenMinMax used for generating inlined version Math.min/max
has been updated to make use of the conditional move support.
Change-Id: I92c5428e40aa8ff88bd3071619957ac3130efae7
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| 090dd4489eeffb5f10051a5d9c1ed71b0a6bc4b9 |
|
20-Dec-2013 |
Razvan A Lupusoru <razvan.a.lupusoru@intel.com> |
Eliminate redundant x86 compare for GenDivZeroCheck
For x86, the ALU operations on general purpose registers update the flags.
Thus, when generating the zero check for divide/remainder operations, the
compare is not needed.
Change-Id: I07bfdf7d5491d3e3e9d98a932472d7f18d5b46d3
Signed-off-by: Razvan A Lupusoru <razvan.a.lupusoru@intel.com>
|
| 5816ed48bc339c983b40dc493e96b97821ce7966 |
|
27-Nov-2013 |
Vladimir Marko <vmarko@google.com> |
Detect special methods at the end of verification.
This moves special method handling to method inliner
and prepares for eventual inlining of these methods.
Change-Id: I51c51b940fb7bc714e33135cd61be69467861352
|
| e13717e796d338b08ea66f6a7e3470ca44de707f |
|
20-Nov-2013 |
Vladimir Marko <vmarko@google.com> |
Per-DexFile locking for inliner initialization.
And clean up lock and compiler driver naming.
Change-Id: I1562c7f55c4b0174a36007ba6199360da06169ff
|
| 31c2aac7137b69d5622eea09597500731fbee2ef |
|
09-Dec-2013 |
Vladimir Marko <vmarko@google.com> |
Rename ClobberCalleeSave to *Caller*, fix it for x86.
Change-Id: I6a72703a11985e2753fa9b4520c375a164301433
|
| 06606b9c4a1c00154ed15f719ad8ea994e54ee8e |
|
02-Dec-2013 |
Vladimir Marko <vmarko@google.com> |
Performance improvement for mapping table creation.
Avoid the raw mapping tables altogether.
Change-Id: I6d1c786325d369e899a75f15701edbafdd14363f
|
| 70b797d998f2a28e39f7d6ffc8a07c9cbc47da14 |
|
03-Dec-2013 |
Vladimir Marko <vmarko@google.com> |
Unsafe.compareAndSwapLong() intrinsic for x86.
Change-Id: Idbc5371a62dfdd84485a657d4548990519200205
|
| 1e6cb63d77090ddc6aa19c755d7066f66e9ff87e |
|
28-Nov-2013 |
Vladimir Marko <vmarko@google.com> |
Delta-encoding of mapping tables.
Both PC offsets and dalvik offsets are delta-encoded. Since
PC offsets are increasing, the deltas are then compressed as
unsigned LEB128. Dalvik offsets are not monotonic, so their
deltas are compressed as signed LEB128.
This reduces the size of the mapping tables by about 30%
on average, 25% from the PC offset and 5% from the dalvik
offset delta encoding.
Bug: 9437697
Change-Id: I600ab9c22dec178088d4947a811cca3bc8bd4cf4
|
| 3e5af82ae1a2cd69b7b045ac008ac3b394d17f41 |
|
21-Nov-2013 |
Vladimir Marko <vmarko@google.com> |
Intrinsic Unsafe.CompareAndSwapLong() for ARM.
(cherry picked from cb53fcd79b1a5ce608208ec454b5c19f64aaba37)
Change-Id: Iadd3cc8b4ed390670463b80f8efd579ce6ece226
|
| 1c282e2b9a9b432e132b2c332f861cad9feb4a73 |
|
21-Nov-2013 |
Vladimir Marko <vmarko@google.com> |
Refactor intrinsic CAS, prepare for 64-bit version.
Bug: 11391018
Change-Id: Ic0f740e0cd0eb47f2c915f81be02f52f7721f8a3
|
| 5c96e6b4dc354a7439b211b93462fbe8edea5e57 |
|
14-Nov-2013 |
Vladimir Marko <vmarko@google.com> |
Rewrite intrinsics detection.
Intrinsic methods should be treated as a special case of
inline methods. They should be detected early and used to
guide other optimizations. This CL rewrites the intrinsics
detection so that it can be moved to any compilation phase.
Change-Id: I4424a6a869bd98b9c478953c9e3bcaf1c6de2b33
|
| e508a2090b19fe705fbc6b99d76474037a74bbfb |
|
04-Nov-2013 |
Vladimir Marko <vmarko@google.com> |
Fix unaligned Memory peek/poke intrinsics.
Change-Id: Id454464d0b28aa37f5239f1c6589ceb0b3bbbdea
|
| 65636e5de2375839e29e3e19ee7a7db737901cf0 |
|
24-Oct-2013 |
Vladimir Marko <vmarko@google.com> |
Add intrinsics for Memory peek/poke.
Add intrinsics for single memory access (non-array)
peek/poke methods in libcore.io.Memory.
Change-Id: I5d66a5b14ea89875d8afb8252eb293f7d637b83f
|
| 6bdf1fff5f841f3997d4b488f00647f7aa2cdaa3 |
|
29-Oct-2013 |
Vladimir Marko <vmarko@google.com> |
Add intrinsics for {Short,Int,Long}.reverseBytes().
Change-Id: I34a2ec642f59fc4ff18aed59769a9e8d7e361098
|
| 0d82948094d9a198e01aa95f64012bdedd5b6fc9 |
|
12-Oct-2013 |
buzbee <buzbee@google.com> |
64-bit prep
Preparation for 64-bit roll.
o Eliminated storing pointers in 32-bit int slots in LIR.
o General size reductions of common structures to reduce impact
of doubled pointer sizes:
- BasicBlock struct was 72 bytes, now is 48.
- MIR struct was 72 bytes, now is 64.
- RegLocation was 12 bytes, now is 8.
o Generally replaced uses of BasicBlock* pointers with 16-bit Ids.
o Replaced several doubly-linked lists with singly-linked to save
one stored pointer per node.
o We had quite a few uses of uintptr_t's that were a holdover from
the JIT (which used pointers to mapped dex & actual code cache
addresses rather than trace-relative offsets). Replaced those with
uint32_t's.
o Clean up handling of embedded data for switch tables and array data.
o Miscellaneous cleanup.
I anticipate one or two additional CLs to reduce the size of MIR and LIR
structs.
Change-Id: I58e426d3f8e5efe64c1146b2823453da99451230
|
| 409fe94ad529d9334587be80b9f6a3d166805508 |
|
11-Oct-2013 |
buzbee <buzbee@google.com> |
Quick assembler fix
This CL re-instates the select pattern optimization disabled by
CL 374310, and fixes the underlying problem: improper handling of
the kPseudoBarrier LIR opcode. The bug was introduced in the
recent assembler restructuring. In short, LIR pseudo opcodes (which
have values < 0), should always have size 0 - and thus cause no
bits to be emitted during assembly. In this case, bad logic caused
us to set the size of a kPseudoBarrier opcode via lookup through the
EncodingMap.
Because all pseudo ops are < 0, this meant we did an array underflow
load, picking up whatever garbage was located before the EncodingMap.
This explains why this error showed up recently - we'd previuosly just
gotten a lucky layout.
This CL corrects the faulty logic, and adds DCHECKs to uses of
the EncodingMap to ensure that we don't try to access w/ a
pseudo op. Additionally, the existing is_pseudo_op() macro is
replaced with IsPseudoLirOp(), named similar to the existing
IsPseudoMirOp().
Change-Id: I46761a0275a923d85b545664cadf052e1ab120dc
|
| a9a8254c920ce8e22210abfc16c9842ce0aea28f |
|
04-Oct-2013 |
Ian Rogers <irogers@google.com> |
Improve quick codegen for aput-object.
1) don't type check known null.
2) if we know types in verify don't check at runtime.
3) if we're runtime checking then move all the code out-of-line.
Also, don't set up a callee-save frame for check-cast, do an instance-of test
then throw an exception if that fails.
Tidy quick entry point of Ldivmod to Lmod which it is on x86 and mips.
Fix monitor-enter/exit NPE for MIPS.
Fix benign bug in mirror::Class::CannotBeAssignedFromOtherTypes, a byte[]
cannot be assigned to from other types.
Change-Id: I9cb3859ec70cca71ed79331ec8df5bec969d6745
|
| d9c4fc94fa618617f94e1de9af5f034549100753 |
|
02-Oct-2013 |
Ian Rogers <irogers@google.com> |
Inflate contended lock word by suspending owner.
Bug 6961405.
Don't inflate monitors for Notify and NotifyAll.
Tidy lock word, handle recursive lock case alongside unlocked case and move
assembly out of line (except for ARM quick). Also handle null in out-of-line
assembly as the test is quick and the enter/exit code is already a safepoint.
To gain ownership of a monitor on behalf of another thread, monitor contenders
must not hold the monitor_lock_, so they wait on a condition variable.
Reduce size of per mutex contention log.
Be consistent in calling thin lock thread ids just thread ids.
Fix potential thread death races caused by the use of FindThreadByThreadId,
make it invariant that returned threads are either self or suspended now.
Code size reduction on ARM boot.oat 0.2%.
Old nexus 7 speedup 0.25%, new nexus 7 speedup 1.4%, nexus 10 speedup 2.24%,
nexus 4 speedup 2.09% on DeltaBlue.
Change-Id: Id52558b914f160d9c8578fdd7fc8199a9598576a
|
| b48819db07f9a0992a72173380c24249d7fc648a |
|
15-Sep-2013 |
buzbee <buzbee@google.com> |
Compile-time tuning: assembly phase
Not as much compile-time gain from reworking the assembly phase as I'd
hoped, but still worthwhile. Should see ~2% improvement thanks to
the assembly rework. On the other hand, expect some huge gains for some
application thanks to better detection of large machine-generated init
methods. Thinkfree shows a 25% improvement.
The major assembly change was to establish thread the LIR nodes that
require fixup into a fixup chain. Only those are processed during the
final assembly pass(es). This doesn't help for methods which only
require a single pass to assemble, but does speed up the larger methods
which required multiple assembly passes.
Also replaced the block_map_ basic block lookup table (which contained
space for a BasicBlock* for each dex instruction unit) with a block id
map - cutting its space requirements by half in a 32-bit pointer
environment.
Changes:
o Reduce size of LIR struct by 12.5% (one of the big memory users)
o Repurpose the use/def portion of the LIR after optimization complete.
o Encode instruction bits to LIR
o Thread LIR nodes requiring pc fixup
o Change follow-on assembly passes to only consider fixup LIRs
o Switch on pc-rel fixup kind
o Fast-path for small methods - single pass assembly
o Avoid using cb[n]z for null checks (almost always exceed displacement)
o Improve detection of large initialization methods.
o Rework def/use flag setup.
o Remove a sequential search from FindBlock using lookup table of 16-bit
block ids rather than full block pointers.
o Eliminate pcRelFixup and use fixup kind instead.
o Add check for 16-bit overflow on dex offset.
Change-Id: I4c6615f83fed46f84629ad6cfe4237205a9562b4
|
| d91d6d6a80748f277fd938a412211e5af28913b1 |
|
26-Sep-2013 |
Ian Rogers <irogers@google.com> |
Introduce Signature type to avoid string comparisons.
Method resolution currently creates strings to then compare with strings formed
from methods in other dex files. The temporary strings are purely created for
the sake of comparisons. This change creates a new Signature type that
represents a method signature but not as a string. This type supports
comparisons and so can be used when searching for methods in resolution.
With this change malloc is no longer the hottest method during dex2oat (now its
memset) and allocations during verification have been reduced. The verifier is
commonly what is populating the dex cache for methods and fields not declared
in the dex file itself.
Change-Id: I5ef0542823fbcae868aaa4a2457e8da7df0e9dae
|
| c729a6b936d59562bd9fb830a595d9ff65dfd129 |
|
15-Sep-2013 |
buzbee <buzbee@google.com> |
Improve promotion of double-precision regs
Minor rework of the double allocation mechanism to more explicitly
manage the allocation of preserved floating point single pairs as
doubles.
Change-Id: Id9db4b0e86e5ef54a5db587f367e00efdf7e98d6
|
| bd663de599b16229085759366c56e2ed5a1dc7ec |
|
11-Sep-2013 |
buzbee <buzbee@google.com> |
Compile-time tuning: register/bb utilities
This CL yeilds about a 4% improvement in the compilation phase
of dex2oat (single-threaded; multi-threaded compilation is
more difficult to accurately measure). The register utilities
could stand to be completely rewritten, but this gets most of the
easy benefit.
Next up: the assembly phase.
Change-Id: Ife5a474e9b1a6d9e501e888dda6749d34eb77e96
|
| 252254b130067cd7a5071865e793966871ae0246 |
|
09-Sep-2013 |
buzbee <buzbee@google.com> |
More Quick compile-time tuning: labels & branches
This CL represents a roughly 3.5% performance improvement for the
compile phase of dex2oat. Move of the gain comes from avoiding
the generation of dex boundary LIR labels unless a debug listing
is requested. The other significant change is moving from a basic block
ending branch model of "always generate a fall-through branch, and then
delete it if we can" to a "only generate a fall-through branch if we need
it" model.
The data motivating these changes follow. Note that two area of
potentially attractive gain remain: restructing the assembler model and
reworking the register handling utilities. These will be addressed
in subsequent CLs.
--- data follows
The Quick compiler's assembler has shown up on profile reports a bit
more than seems reasonable. We've tried a few quick fixes to apparently
hot portions of the code, but without much gain. So, I've been looking at
the assembly process at a somewhat higher level. There look to be several
potentially good opportunities.
First, an analysis of the makeup of the LIR graph showed a surprisingly
high proportion of LIR pseudo ops. Using the boot classpath as a basis,
we get:
32.8% of all LIR nodes are pseudo ops.
10.4% are LIR instructions which require pc-relative fixups.
11.8% are LIR instructions that have been nop'd by the various
optimization passes.
Looking only at the LIR pseudo ops, we get:
kPseudoDalvikByteCodeBoundary 43.46%
kPseudoNormalBlockLabel 21.14%
kPseudoSafepointPC 20.20%
kPseudoThrowTarget 6.94%
kPseudoTarget 3.03%
kPseudoSuspendTarget 1.95%
kPseudoMethodExit 1.26%
kPseudoMethodEntry 1.26%
kPseudoExportedPC 0.37%
kPseudoCaseLabel 0.30%
kPseudoBarrier 0.07%
kPseudoIntrinsicRetry 0.02%
Total LIR count: 10167292
The standout here is the Dalvik opcode boundary marker. This is just a
label inserted at the beginning of the codegen for each Dalvik bytecode.
If we're also doing a verbose listing, this is also where we hang the
pretty-print disassembly string. However, this label was also
being used as a convenient way to find the target of switch case
statements (and, I think at one point was used in the Mir->GBC conversion
process).
This CL moves the use of kPseudoDalvikByteCodeBoundary labels to only
verbose listing runs, and replaces the codegen uses of the label with
the kPseudoNormalBlockLabel attached to the basic block that contains the
switch case target. Great savings here - 14.3% reduction in the number of
LIR nodes needed. After this CL, our LIR pseudo proportions drop to 21.6%
of all LIR. That's still a lot, but much better. Possible further
improvements via combining normal labels with kPseudoSafepointPC labels
where appropriate, and also perhaps reduce memory usage by using a
short-hand form for labels rather than a full LIR node. Also, many
of the basic block labels are no longer branch targets by the time
we get to assembly - cheaper to delete, or just ingore?
Here's the "after" LIR pseudo op breakdown:
kPseudoNormalBlockLabel 37.39%
kPseudoSafepointPC 35.72%
kPseudoThrowTarget 12.28%
kPseudoTarget 5.36%
kPseudoSuspendTarget 3.45%
kPseudoMethodEntry 2.24%
kPseudoMethodExit 2.22%
kPseudoExportedPC 0.65%
kPseudoCaseLabel 0.53%
kPseudoBarrier 0.12%
kPseudoIntrinsicRetry 0.04%
Total LIR count: 5748232
Not done in this CL, but it will be worth experimenting with actually
deleting LIR nodes from the graph when they are optimized away, rather
than just setting the NOP bit. Keeping them around is invaluable
during debugging - but when not debugging it may pay off if the cost of
node removal is less than the cost of traversing through dead nodes
in subsequent passes.
Next up (and partially in this CL - but mostly to be done in follow-on
CLs) is the overall assembly process. Inherited from the trace JIT,
the Quick compiler has a fairly simple-minded approach to instruction
assembly. First, a pass is made over the LIR list to assign offsets
to each instruction. Then, the assembly pass is made - which generates
the actual machine instruction bit patterns and pushes the instruction
data into the code_buffer. However, the code generator takes the "always
optimistic" approach to instruction selection and emits the shortest
instruction. If, during assembly, we find that a branch or load doesn't
reach, that short-form instruction is replaces with a longer sequence.
Of course, this invalidates the previously-computed offset calculations.
Assembly thus is an iterative process: compute offsets and then assemble
until we survive an assembly pass without invalidation. This seems
like a likely candidate for improvement. First, I analyzed the
number of retries required, and the reason for invalidation over the
boot classpath load.
The results: more than half of methods don't require a retry, and
very few require more than 1 extra pass:
5 or more: 6 of 96334
4 or more: 22 of 96334
3 or more: 140 of 96334
2 or more: 1794 of 96334 - 2%
1 or more: 40911 of 96334 - 40%
0 retries: 55423 of 96334 - 58%
The interesting group here is the one that requires 1 retry. Looking
at the reason, we see three typical reasons:
1. A cbnz/cbz doesn't reach (only 7 bits of offset)
2. A 16-bit Thumb1 unconditional branch doesn't reach.
3. An unconditional branch which branches to the next instruction
is encountered, and deleted.
The first 2 cases are the cost of the optimistic strategy - nothing
much to change there. However, the interesting case is #3 - dead
branch elimination. A further analysis of the single retry group showed
that 42% of the methods (16305) that required a single retry did so
*only* because of dead branch elimination. The big question here is
why so many dead branches survive to the assembly stage. We have
a dead branch elimination pass which is supposed to catch these - perhaps
it's not working correctly, should be moved later in the optimization
process, or perhaps run multiple times.
Other things to consider:
o Combine the offset generation pass with the assembly pass. Skip
pc-relative fixup assembly (other than assigning offset), but push
LIR* for them into work list. Following the main pass, zip through
the work list and assemble the pc-relative instructions (now that we
know the offsets). This would significantly cut back on traversal
costs.
o Store the assembled bits into both the code buffer and the LIR.
In the event we have to retry, only the pc-relative instructions
would need to be assembled, and we'd finish with a pass over the
LIR just to dumb the bits into the code buffer.
Change-Id: I50029d216fa14f273f02b6f1c8b6a0dde5a7d6a6
|
| 56c717860df2d71d66fb77aa77f29dd346e559d3 |
|
06-Sep-2013 |
buzbee <buzbee@google.com> |
Compile-time tuning
Specialized the dataflow iterators and did a few other minor tweaks.
Showing ~5% compile-time improvement in a single-threaded environment;
less in multi-threaded (presumably because we're blocked by something
else).
Change-Id: I2e2ed58d881414b9fc97e04cd0623e188259afd2
|
| 9b297bfc588c7d38efd12a6f38cd2710fc513ee3 |
|
06-Sep-2013 |
Ian Rogers <irogers@google.com> |
Refactor CompilerDriver::Compute..FieldInfo
Don't use non-const reference arguments.
Move ins before outs.
Change-Id: I7b251156388d8f07513b3da62ebfd29e5fd9ff76
|
| 11b63d13f0a3be0f74390b66b58614a37f9aa6c1 |
|
27-Aug-2013 |
buzbee <buzbee@google.com> |
Quick compiler: division by literal fix
The constant propagation optimization pass attempts to identify
constants in Dalvik virtual registers and handle them more efficiently.
The use of small constants in divison, though, was handled incorrectly
in that the high level code correctly detected the use of a constant,
but the actual code generation routine was only expecting the use of
a special constant form opcode.
see b/10503566
Change-Id: I88aa4d2eafebb2b1af1a1e88049f1845aefae261
|
| 96faf5b363d922ae91cf25404dee0e87c740c7c5 |
|
10-Aug-2013 |
Ian Rogers <irogers@google.com> |
Uleb128 compression of vmap and mapping table.
Bug 9437697.
Change-Id: I30bcb97d12cd8b46d3b2cdcbdd358f08fbb9947a
(cherry picked from commit 1809a72a66d245ae598582d658b93a24ac3bf01e)
|
| 468532ea115657709bc32ee498e701a4c71762d4 |
|
05-Aug-2013 |
Ian Rogers <irogers@google.com> |
Entry point clean up.
Create set of entry points needed for image methods to avoid fix-up at load time:
- interpreter - bridge to interpreter, bridge to compiled code
- jni - dlsym lookup
- quick - resolution and bridge to interpreter
- portable - resolution and bridge to interpreter
Fix JNI work around to use JNI work around argument rewriting code that'd been
accidentally disabled.
Remove abstact method error stub, use interpreter bridge instead.
Consolidate trampoline (previously stub) generation in generic helper.
Simplify trampolines to jump directly into assembly code, keeps stack crawlable.
Dex: replace use of int with ThreadOffset for values that are thread offsets.
Tidy entry point routines between interpreter, jni, quick and portable.
Change-Id: I52a7c2bbb1b7e0ff8a3c3100b774212309d0828e
(cherry picked from commit 848871b4d8481229c32e0d048a9856e5a9a17ef9)
|
| 1809a72a66d245ae598582d658b93a24ac3bf01e |
|
10-Aug-2013 |
Ian Rogers <irogers@google.com> |
Uleb128 compression of vmap and mapping table.
Bug 9437697.
Change-Id: I30bcb97d12cd8b46d3b2cdcbdd358f08fbb9947a
|
| 848871b4d8481229c32e0d048a9856e5a9a17ef9 |
|
05-Aug-2013 |
Ian Rogers <irogers@google.com> |
Entry point clean up.
Create set of entry points needed for image methods to avoid fix-up at load time:
- interpreter - bridge to interpreter, bridge to compiled code
- jni - dlsym lookup
- quick - resolution and bridge to interpreter
- portable - resolution and bridge to interpreter
Fix JNI work around to use JNI work around argument rewriting code that'd been
accidentally disabled.
Remove abstact method error stub, use interpreter bridge instead.
Consolidate trampoline (previously stub) generation in generic helper.
Simplify trampolines to jump directly into assembly code, keeps stack crawlable.
Dex: replace use of int with ThreadOffset for values that are thread offsets.
Tidy entry point routines between interpreter, jni, quick and portable.
Change-Id: I52a7c2bbb1b7e0ff8a3c3100b774212309d0828e
|
| 7934ac288acfb2552bb0b06ec1f61e5820d924a4 |
|
26-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix cpplint whitespace/comments issues
Change-Id: Iae286862c85fb8fd8901eae1204cd6d271d69496
|
| 6f485c62b9cfce3ab71020c646ab9f48d9d29d6d |
|
19-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix cpplint whitespace/indent issues
Change-Id: I7c1647f0c39e1e065ca5820f9b79998691ba40b1
|
| 9b7085a4e7c40e7fa01932ea1647a4a33ac1c585 |
|
19-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix cpplint readability/braces issues
Change-Id: I56b88956510077b0e13aad4caee8898313fab55b
|
| df62950e7a32031b82360c407d46a37b94188fbb |
|
18-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix cpplint whitespace/parens issues
Change-Id: Ifc678d59a8bed24ffddde5a0e543620b17b0aba9
|
| 0cd7ec2dcd8d7ba30bf3ca420b40dac52849876c |
|
18-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix cpplint whitespace/blank_line issues
Change-Id: Ice937e95e23dd622c17054551d4ae4cebd0ef8a2
|
| 2ce745c06271d5223d57dbf08117b20d5b60694a |
|
18-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix cpplint whitespace/braces issues
Change-Id: Ide80939faf8e8690d8842dde8133902ac725ed1a
|
| fc0e3219edc9a5bf81b166e82fd5db2796eb6a0d |
|
17-Jul-2013 |
Brian Carlstrom <bdc@google.com> |
Fix multiple inclusion guards to match new pathnames
Change-Id: Id7735be1d75bc315733b1773fba45c1deb8ace43
|
| 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
|