Pat Gavlin [Wed, 14 Sep 2016 21:22:52 +0000 (14:22 -0700)]
Address more PR feedback.
Pat Gavlin [Wed, 14 Sep 2016 20:18:06 +0000 (13:18 -0700)]
Address PR feedback.
Pat Gavlin [Wed, 14 Sep 2016 02:32:57 +0000 (19:32 -0700)]
Introduce GT_JCC.
This node represents a jump that is conditional upon the value stored in
the target's condition code register. It is only valid in the backend.
No formal modeling of the CCR is performed, so its use must be
constrained such that instructions that def the CCR are not inserted
between the JCC node and the node that it expected to def the CCR.
This is currently only used when lowering compares of long-typed
values for x86.
Mike Danes [Thu, 8 Sep 2016 11:54:14 +0000 (14:54 +0300)]
Implement long compare lowering for x86.
Comparisons between long-typed values on x86 require at least two and
at most three separate branches: one or two to compare the high 32
bits and one to compare the low 32 bits. In essence, each long
compare introduces two additional basic blocks into the flow graph,
but these blocks were not reified until code generation. Furthermore,
code generation of a long comparison used by a JTRUE was deferred
until the JTRUE itself without marking the inputs to the compare as
live at the JTRUE. Taken together, these representational issues
caused bugs like the one seen in the the following assembly:
33DB xor ebx, ebx
BE03000000 mov esi, 3
33FF xor edi, edi
8B75F0 mov esi, dword ptr [ebp-10H] ; these 2 are reloads inserted by LSRA
8B7DEC mov edi, dword ptr [ebp-14H] ; between the compare and the branch
3BDF cmp ebx, edi
72D6 jb SHORT G_M51005_IG03
7704 ja SHORT G_M51005_IG04
3BC6 cmp eax, esi
72D0 jb SHORT G_M51005_IG03
The reloads that LSRA has inserted have killed the registers assigned
to the inputs to the compare, thus causing the compare instructions
to read unexpected values (GH #7038).
Although a number of alternatives were discussed, the best solution
seems to be to expose the control flow via the flow graph rather than
leaving it until code generation; this commit implements that approach.
Peter Kukol [Tue, 13 Sep 2016 04:35:54 +0000 (22:35 -0600)]
Clean up COMPlus_JitMemStats / MEASURE_MEM_ALLOC (#7119)
Clean up COMPlus_JitMemStats / MEASURE_MEM_ALLOC (all of it is off by default for non-DEBUG builds).
Michelle McDaniel [Tue, 13 Sep 2016 04:18:59 +0000 (21:18 -0700)]
Merge pull request #7105 from adiaaida/umodLong
Implement DecomposeUMod
Michelle McDaniel [Thu, 8 Sep 2016 15:52:37 +0000 (08:52 -0700)]
Implement DecomposeUMod
This change implements DecomposeUMod for x86 RyuJIT. The only GT_UMOD
nodes that make it to decompose are ones where op2 is a cast from a
constant int to long. Because op2 is an int, we can guarantee that the
result will be an int. Therefore, in decompose, we change the type of the
GT_UMOD to be TYP_INT, and replace op2 with its lo part. We set the high
part of the GT_LONG to be 0, since the divisor < 0x3fffffff. In lower, we need
to make sure that loOp1 is in RAX and hiOp1 is in RDX, which is where
idiv expects them to be. We also increase the number of sources since
there are now three sources. In codegen, we need to make sure that the hi
and lo parts of the dividend are in the correct registers for idiv, then
we can just use the normal logic for GT_UMOD.
Pat Gavlin [Tue, 13 Sep 2016 00:34:49 +0000 (17:34 -0700)]
Merge pull request #7126 from mikedn/x86-cast-long-int-overflow
Implement long to int casts with overflow for x86
Bruce Forstall [Mon, 12 Sep 2016 22:36:29 +0000 (15:36 -0700)]
Merge pull request #7148 from BruceForstall/Fix7127
Restrict x86 legacy JIT struct promotion
Pat Gavlin [Mon, 12 Sep 2016 22:28:26 +0000 (15:28 -0700)]
Merge pull request #7133 from mikedn/x86-cast-long-float
Implement long to float cast for x86
Vance Morrison [Mon, 12 Sep 2016 22:16:59 +0000 (15:16 -0700)]
Merge pull request #7102 from vancem/ETWGCDumpFix.9-7-16
Fix Issue dumping GC heap with ETW more than once
Bruce Forstall [Mon, 12 Sep 2016 21:19:12 +0000 (14:19 -0700)]
Restrict x86 legacy JIT struct promotion
Don't allow struct promotion for custom layout structs. RyuJIT started
supporting this, but the legacy x86 JIT doesn't properly push such promoted
structs as arguments for calls.
Fixes the assert in #7127, but the test fails with a crash in a later test case.
Jan Kotas [Mon, 12 Sep 2016 20:39:11 +0000 (13:39 -0700)]
Merge pull request #7136 from benaadams/enviro
Updated formatted GetResourceString code reduction
Jose Perez Rodriguez [Mon, 12 Sep 2016 18:42:46 +0000 (11:42 -0700)]
Merge pull request #7142 from joperezr/FillOutMissingStringMembers
Adding missing String members to coreclr
Davis Goodin [Mon, 12 Sep 2016 18:15:48 +0000 (13:15 -0500)]
Merge pull request #6791 from dagood/upgrade-xunit-runner
Upgrade xunit runner to 2.2.0-beta2-build3300
Jose Perez Rodriguez [Mon, 12 Sep 2016 17:10:22 +0000 (10:10 -0700)]
Removing one String member to avoid depending on ArgIterator
Jose Perez Rodriguez [Mon, 12 Sep 2016 16:44:39 +0000 (09:44 -0700)]
Adding missing String members to coreclr
Mike Danes [Sun, 11 Sep 2016 12:49:12 +0000 (15:49 +0300)]
Implement long to float cast for x86
Convert long to float/double casts to helper calls. Despite the call this version is 2x faster than JIT32's FILD implementation which suffers a significant store forwarding stall penalty.
Ben Adams [Sun, 11 Sep 2016 21:26:19 +0000 (22:26 +0100)]
Fix Environment no-inlining for mscorlib
James Ko [Wed, 31 Aug 2016 15:12:13 +0000 (11:12 -0400)]
Reduce code bloat around formatted GetResourceString calls
Peter Kukol [Sat, 10 Sep 2016 00:27:19 +0000 (18:27 -0600)]
Merge pull request #7067 from pkukol/timing_bit_rot
Fix some timing code bit rot, plus minor cleanup.
Pat Gavlin [Fri, 9 Sep 2016 23:22:07 +0000 (16:22 -0700)]
Merge pull request #7125 from pgavlin/ModelAssignByRefKillSet
Model the kill set for ASSIGN_BYREF and stop generating movsq on x86.
Bruce Forstall [Fri, 9 Sep 2016 22:02:29 +0000 (15:02 -0700)]
Merge pull request #7122 from BruceForstall/Fix7092
Fix #7092
Bruce Forstall [Fri, 9 Sep 2016 21:59:09 +0000 (14:59 -0700)]
Merge pull request #7120 from BruceForstall/Fix7093
Fix #7093
Bruce Forstall [Fri, 9 Sep 2016 21:58:08 +0000 (14:58 -0700)]
Merge pull request #7114 from BruceForstall/Fix7100
Fix #7100
Mike Danes [Fri, 9 Sep 2016 20:44:29 +0000 (23:44 +0300)]
Implement long to int casts with overflow for x86
Bruce Forstall [Fri, 9 Sep 2016 17:47:45 +0000 (10:47 -0700)]
Fix #7092
The problem is genCodeForInitBlkUnroll() is trying to generate:
```
mov byte ptr [edi], si
```
which is not legal: ebp/esi/edi/esp don't have byte register versions on x86.
This change requires the source register to be an x86 byteable register in
this case.
Note that legacy JIT generates stosb for this test.
Pat Gavlin [Fri, 9 Sep 2016 19:26:44 +0000 (12:26 -0700)]
Do not use movsq on x86.
This instruction is only available when targeting amd64.
Pat Gavlin [Fri, 9 Sep 2016 18:04:19 +0000 (11:04 -0700)]
Model the kill set for ASSIGN_BYREF on x86.
This helper kills esi, edi, and ecx.
Sivarv [Fri, 9 Sep 2016 17:16:16 +0000 (10:16 -0700)]
Merge pull request #7111 from sivarv/jitstressregs2Fix
Fix to #7087 - x86 RyuJIT JitstressRegs=2 assert failure
Kyungwoo Lee [Fri, 9 Sep 2016 16:32:39 +0000 (10:32 -0600)]
Merge pull request #7108 from kyulee1/igmeta
ILDAMS: Ignore Metadata flag for CoreCLR
Sean Gillespie [Fri, 9 Sep 2016 16:32:24 +0000 (09:32 -0700)]
Merge pull request #7121 from swgillespie/request-build-break
Only reference IGCHeap::gcHeapType when FEATURE_SVR_GC is defined
Sean Gillespie [Fri, 9 Sep 2016 13:53:48 +0000 (06:53 -0700)]
Only reference IGCHeap::gcHeapType when FEATURE_SVR_GC is defined
Jan Vorlicek [Fri, 9 Sep 2016 08:57:45 +0000 (10:57 +0200)]
Fix enregistered max size constants on Unix AMD64 (#7117)
This change fixes ENREGISTERED_PARAMTYPE_MAXSIZE and ENREGISTERED_RETURNTYPE_INTEGER_MAXSIZE
constants that were not updated properly for the Unix AMD64 ABI. The values are used
for Unix/AMD64 for marshalling and typedbyref parameters and our tests probably don't
execute those code paths with structs larger than 8 bytes.
Jan Kotas [Fri, 9 Sep 2016 02:35:16 +0000 (19:35 -0700)]
Merge pull request #7118 from dotnet/revert-6889-StaticField
Revert "Fix IMGREL32 static field addr value-num blindspot"
Bruce Forstall [Fri, 9 Sep 2016 01:09:34 +0000 (18:09 -0700)]
Fix #7093
This is an assert due to how STRESS_64RSLT_MUL is implemented. This
stress mode converts:
```
/--* lclVar int V01 loc0
* * int
\--* lclVar int V01 loc0
```
to:
```
* cast int <- long
| /--* cast long <- int
| | \--* lclVar int V01 loc0
\--* * long
\--* cast long <- long
\--* nop long
\--* lclVar int V01 loc0
```
Thus, the long 'nop' node is above an 'int' operand node. This led to an assert
in genCodeForTreeLng() that the lclVar was type long, which is isn't.
I added yet another cast under the 'nop' to fix this typing problem:
```
* cast int <- long
| /--* cast long <- int
| | \--* lclVar int V01 loc0
\--* * long
\--* cast long <- long
\--* nop long
\--* cast long <- int
\--* lclVar int V01 loc0
```
Jan Kotas [Fri, 9 Sep 2016 00:58:55 +0000 (17:58 -0700)]
Revert "Fix IMGREL32 static field addr value-num blindspot"
Sean Gillespie [Fri, 9 Sep 2016 00:45:58 +0000 (17:45 -0700)]
Merge pull request #7116 from dotnet-bot/from-tfs
Merge changes from TFS
dotnet-bot [Fri, 9 Sep 2016 00:24:16 +0000 (17:24 -0700)]
Fix a build break introduced by GH#6764
[tfs-changeset: 1626397]
Bruce Forstall [Thu, 8 Sep 2016 23:42:22 +0000 (16:42 -0700)]
Fix #7100
This issue is the following assert:
```
Assert failure '!emitComp->opts.compReloc || memBase->IsIconHandle()'
```
while ngen'ing mscorlib on desktop.
We're trying to generate
```
cmp ebx, dword ptr [0000H]
```
(due to inlining). But, emitInsBinary() doesn't expect a zero address that
is not a relocatable handle.
Simply change the assert to allow this. The code then generates the correct
zero base address.
Rahul Kumar [Thu, 8 Sep 2016 23:27:42 +0000 (16:27 -0700)]
Merge pull request #7112 from rahku/appdomain
Add AppContext.UnhandledException to ref file
Rahul Kumar [Thu, 8 Sep 2016 23:17:06 +0000 (16:17 -0700)]
Add AppContext.UnhandledException to ref file
sivarv [Thu, 8 Sep 2016 22:39:08 +0000 (15:39 -0700)]
Fix to #7087
Eugene Rozenfeld [Thu, 8 Sep 2016 22:28:47 +0000 (15:28 -0700)]
Merge pull request #7071 from erozenfeld/StackOverflow
Make GT_LIST processing non-recursive to avoid StackOverflow.
Kyungwoo Lee [Thu, 8 Sep 2016 19:59:39 +0000 (12:59 -0700)]
ILDASM: Ignore Metadata flag for CoreCLR
Full metadata emission is disabled under coreclr.
So is ILDASM instead of failing during the run.
Peter Kukol [Wed, 7 Sep 2016 15:58:44 +0000 (08:58 -0700)]
Fix the fixes.
Peter Kukol [Tue, 6 Sep 2016 18:28:30 +0000 (11:28 -0700)]
Fix some timing code bit rot, plus minor cleanup.
Pat Gavlin [Thu, 8 Sep 2016 20:24:00 +0000 (13:24 -0700)]
Merge pull request #7107 from pgavlin/gh7089
Fix #7089.
Sean Gillespie [Thu, 8 Sep 2016 20:06:21 +0000 (13:06 -0700)]
Merge pull request #6764 from swgillespie/gc-interface-3
Move the GC behind an interface and use that interface in the VM
Sean Gillespie [Thu, 8 Sep 2016 18:27:24 +0000 (11:27 -0700)]
Introduce an interface separating the GC and the VM,
modifying the VM to utilize this interface.
Introduce an interface separating the GC and the rest of the VM
Remove static members of both IGCHeap and IGCHeapInternal and move the management of the singular GC heap to the VM.
Rename uses of IGCHeap in the VM to GCHeapHolder, as well as other misc. renames throughout the VM and GC.
Split each interface function into categories, document them, use consistent formatting across the interface
Undo some accidental find/replace collateral damage
Remove all ifdefs from the GC interface
Deduplicate function declarations between IGCHeap and IGCHeapInternal, expose AllocAlign8 through the interface and the reference to alloc_context to repair the ARM build
Paper cut: false -> nullptr
Repair the ARM and x86 builds
Rename GCHeapHolder -> GCHeapUtilities and address documentation feedback
Rebase against master
Rename gcholder.h/cpp -> gcheaputilities.h/cpp
Fix an uninitialized field on alloc_context causing test failures on clang
Rename the include guard for gcheaputilities.h
Un-breaks SOS by making the following changes:
1) Instructs the DAC to look for IGCHeap::gcHeapType by name,
instead of assuming that it exists near g_pGCHeap,
2) Eliminate all virtual calls on IGCHeap in the DAC, since we cannot
dispatch on an object in another process,
3) Because of 2, expose the number of generations past the GC interface
using a static variable on IGCHeap that the DAC can read directly.
repair the Windows build
Pat Gavlin [Thu, 8 Sep 2016 17:52:21 +0000 (10:52 -0700)]
Merge pull request #7106 from pgavlin/DecompTP
Improve decomp throughput.
Pat Gavlin [Thu, 8 Sep 2016 16:36:20 +0000 (09:36 -0700)]
Return false from isRegPairType for RyuJIT.
The RyuJIT backend does not use regpair types, so this method should
simply return false when that backend is in use. In the future we
should consider removing this method (and other regpair-related code).
Fixes #7089.
Rahul Kumar [Thu, 8 Sep 2016 16:44:01 +0000 (09:44 -0700)]
Merge pull request #6995 from rahku/appdomain
Exposes runtime functionality to Appdomain implementation in corefx
Pat Gavlin [Thu, 8 Sep 2016 16:31:08 +0000 (09:31 -0700)]
Merge pull request #4986 from mikedn/x86-cast-int-long
Implement integer to long/ulong casts for x86
Pat Gavlin [Thu, 8 Sep 2016 14:27:36 +0000 (07:27 -0700)]
Improve decomp throughput.
Don't find uses unless necessary. This saves about 50 milliseconds while
crossgen'ing mscorlib.
Rahul Kumar [Tue, 30 Aug 2016 21:56:11 +0000 (14:56 -0700)]
Exposes runtime functionality to Appdomain implementation in corefx
Pat Gavlin [Thu, 8 Sep 2016 15:20:47 +0000 (08:20 -0700)]
Merge pull request #7011 from svick/earlyprop-unused-var
Fix BBF_HAS_NULLCHECK test
Mike Danes [Thu, 8 Sep 2016 12:14:32 +0000 (15:14 +0300)]
Add a NYI for a long compare codegen issue
Tracked by #7038
Mike Danes [Sun, 15 May 2016 11:04:13 +0000 (14:04 +0300)]
Implement integer to long/ulong casts for x86
Yufei Huang [Thu, 8 Sep 2016 07:07:54 +0000 (15:07 +0800)]
Add constraint to Enum.Parse<TEnum> (#7104)
Joseph Tremoulet [Thu, 8 Sep 2016 05:04:25 +0000 (01:04 -0400)]
Merge pull request #6889 from JosephTremoulet/StaticField
Fix IMGREL32 static field addr value-num blindspot
Joseph Tremoulet [Thu, 8 Sep 2016 05:04:07 +0000 (01:04 -0400)]
Merge pull request #6754 from hseok-oh/castd2ul
disable gtFoldExprConst for conversion from negative double to unsigned long
Bart J.F. De Smet [Thu, 8 Sep 2016 01:37:45 +0000 (18:37 -0700)]
Make TryEnsureSufficientExecutionStack public (#7077)
Jan Vorlicek [Thu, 8 Sep 2016 01:37:07 +0000 (03:37 +0200)]
Fix OSX build break in PAL tests (#7101)
This fixes a build break in PAL tests introduced by a recent checkin
that converted PAL tests to C++.
Vance Morrison [Wed, 7 Sep 2016 23:32:54 +0000 (16:32 -0700)]
Fix issue where Type Names are not resolved if a heap dump is done twice.
The issue is that a table that remembers which types have been dumped is only
flushed when the 'Types' ETW keyword is disabled, but we only check this when
and ENABLE command is executed. We should do so on DISABLE a well
Pat Gavlin [Wed, 7 Sep 2016 21:56:39 +0000 (14:56 -0700)]
Merge pull request #7023 from pgavlin/X87MathIntrinsics
Stop treating sine/cosine/round as target intrinsics on x86.
Pat Gavlin [Wed, 7 Sep 2016 21:45:36 +0000 (14:45 -0700)]
Link #7097 to the relevant code in ::IsTargetIntrinsic.
kchoi [Wed, 7 Sep 2016 21:12:22 +0000 (14:12 -0700)]
Fix strict aliasing violation from conditional typedef of wchar_t by building entire project as C++ for Unix (#6801)
Enable building CoreCLR as C++ project on Unix
This series of patches fixes the strict aliasing violation from
the conditional typedef of wchar_t in src/pal/inc/pal_char16.h:40
* rename c files to cpp
* modify all cmake files to change .c files to .cpp
* apply c++ linkage to templates
smile21prc [Wed, 7 Sep 2016 20:51:35 +0000 (13:51 -0700)]
Merge pull request #7016 from smile21prc/coreclr-perf
Create coreclr perf jobs for Linux.
Addressed all the comments, and verified coreclr perf run for Linux with latest coreclr/corefx bits passed:
http://dotnet-ci2.cloudapp.net/job/smile21prc_coreclr/job/coreclr-perf/job/perf_Linux/28/consoleFull
Pat Gavlin [Wed, 7 Sep 2016 20:19:31 +0000 (13:19 -0700)]
Merge pull request #7098 from pgavlin/FixTestOutputPath
Fix the output path when building a single test.
Smile Wei [Tue, 30 Aug 2016 19:31:25 +0000 (12:31 -0700)]
Create coreclr perf jobs for Linux.
Pat Gavlin [Wed, 7 Sep 2016 18:09:40 +0000 (11:09 -0700)]
Fix the output path when building a single test.
Add a trailing slash to the base output path.
Pat Gavlin [Wed, 7 Sep 2016 17:51:29 +0000 (10:51 -0700)]
Stop treating sine/cosine/round as target intrinsics on x86.
These intrinsics were supported by JIT32 using the corresponding
x87 instructions. There are a few downsides to doing the same in
RyuJIT, however, because it uses SSE for all other floating-point
math:
- The mix of precisions used by the combination of SSE + x87
instructions matches neither the JIT32 nor the RyuJIT/x64
behavior.
- Using the x87 instructions is more expensive, as the result of
the instruction may need to be moved via memory to an SSE
register.
- The RA would need to be udpated in order to better understand
that each x87 instruction may require a spill temp in order to
retrieve the result.
Issue #7097 tracks investigating a custom calling convention for
the appropriate helper calls if necessasry.
Eugene Rozenfeld [Tue, 6 Sep 2016 23:51:23 +0000 (16:51 -0700)]
Make GT_LIST processing non-recursive to avoid StackOverflow.
We had some internal cases where crossgen failed with StackOverflow exception
when compiling huge methods. In particular, the methods had GT_PHI nodes with
huge number of arguments.
StackOverflow was happening in multiple places. Recent LIR changes eliminated two of those places,
these changes eliminate two more: gtSetEvalOrder and fgDebugCheckFlags (debug only). We already had
gtSetListOrder but it was only used for call arg lists. I made gtSetListOrder non-recursive and also generalized
to handle other GT_LIST nodes.
With that with these changes the huge repros can now be crossgen'd.
I verified no assembly diffs in SuperPMI.
I'm verifying overall throughput effect.
Michelle McDaniel [Wed, 7 Sep 2016 15:55:06 +0000 (08:55 -0700)]
Merge pull request #7065 from adiaaida/multiplyLongMul_Hi
Enable long multiply
Wes Haggard [Wed, 7 Sep 2016 15:54:26 +0000 (08:54 -0700)]
Merge pull request #7081 from yufeih/enum_try_parse
Add Enum.TryParse and Enum.Parse<TEnum>
Ben Adams [Wed, 7 Sep 2016 15:51:49 +0000 (16:51 +0100)]
ConcurrentDictionary<TKey, TValue> Exceptions to ThrowHelper (#7079)
Carol Eidt [Wed, 7 Sep 2016 15:39:42 +0000 (08:39 -0700)]
Merge pull request #7068 from CarolEidt/FixArm64
Fix Arm64 build breaks
Tarek Mahmoud Sayed [Wed, 7 Sep 2016 15:25:01 +0000 (08:25 -0700)]
Merge pull request #7075 from tarekgh/DisableZh-CHS-CHT
Avoid creating zh-CHS/zh-CHT durng culture enumeration
Michelle McDaniel [Wed, 31 Aug 2016 15:14:57 +0000 (08:14 -0700)]
Enable long multiply
Most long multiplies are converted in morph to helper calls. However,
GT_MULs of the form long = (long)int * (long)int are passed through to be
handled in codegen. In decompose, we convert these multiplies to
GT_MUL_HIs to be handled in a similar manner at GT_MULHI. Since mul and
imul take two ints and return a long in edx:eax, we can handle them
similarly to GT_CALLs, where we save them to lclvars if they aren't
already saved. In codegen, we generate a mul (or imul for signed
multiply), which produces the output in edx:eax, and then we store those
locations when we handle the storelclvar.
Hyung-Kyu Choi [Wed, 7 Sep 2016 13:50:44 +0000 (22:50 +0900)]
Exclude smallFrame.sh test for ARM (#7080)
JIT/Methodical/tailcall_v4/smallFrame/smallFrame.sh contains
"tail.call, pop, ret" pattern. The pattern is not allowed in ECMA and
is only supported on x64 for compatibility. (issue #7036)
Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
James Ko [Wed, 7 Sep 2016 13:01:33 +0000 (09:01 -0400)]
Correct the virtual overload of Stream.CopyTo (#7082)
Yufei Huang [Wed, 7 Sep 2016 08:53:59 +0000 (16:53 +0800)]
Add Enum.TryParse and Enum.Parse<TEnum>
Ben Adams [Wed, 7 Sep 2016 05:49:53 +0000 (06:49 +0100)]
Task<T> Exceptions to ThrowHelper (#7076)
Hyeongseok Oh [Wed, 7 Sep 2016 04:26:05 +0000 (13:26 +0900)]
modify comment
Hyeongseok Oh [Wed, 7 Sep 2016 04:18:39 +0000 (13:18 +0900)]
add comment
Tarek Mahmoud Sayed [Wed, 7 Sep 2016 01:33:41 +0000 (18:33 -0700)]
Avoid creating zh-CHS/zh-CHT durng culture enumeration
zh-CHS and zh-CHT are old names that is not really supported in the standards.
when we enabled CultureInfo.GetCultures() there was some code that create cultures
for zh-CHS and zh-CHT. The fix here is disable that otherwise we'll have to enable
a good chunk of legacy code to support it while it never supported by coreclr and in
same time are not names used in standards
James Ko [Wed, 7 Sep 2016 01:19:51 +0000 (21:19 -0400)]
Make Stream.CopyTo virtual (#7070)
Rama krishnan Raghupathy [Wed, 7 Sep 2016 00:20:54 +0000 (17:20 -0700)]
Merge pull request #7022 from lemmaa/prevent-build-output-contamination
[Script] Prevent build output contamination while building test overlay
Joseph Tremoulet [Tue, 6 Sep 2016 19:54:32 +0000 (15:54 -0400)]
Merge pull request #7064 from JosephTremoulet/MakeCse
Fix GTF_ flag collision
Carol Eidt [Tue, 6 Sep 2016 19:39:29 +0000 (12:39 -0700)]
Fix Arm64 build breaks
The `GT_STORE_OBJ` case in codegen was missing a break.
`getStructHandleIfPresent` wasn't correctly handling `GT_ASG'.
The `srcCount` was not being set correctly for block ops with a constant size.
And I lost some code in `fgMakeTmpArgNode` with a merge.
Also, I had introduced a dumping issue in gentree.cpp when I made `GT_OBJ` a `GTK_UNOP` as a result of some PR comments.
Gaurav Khanna [Tue, 6 Sep 2016 19:42:14 +0000 (12:42 -0700)]
Merge pull request #6922 from gkhanna79/WinArm32
Enable Arm32 Test Build
Ben Adams [Tue, 6 Sep 2016 18:14:42 +0000 (19:14 +0100)]
Array<T> exceptions to ThrowHelper (#7059)
Dmitri-Botcharnikov [Tue, 6 Sep 2016 17:27:13 +0000 (21:27 +0400)]
[Linux][LLDB][GDB/JIT] Add support for lldb 'breakpoint set' for jitted code (#6956)
* Add support for lldb 'breakpoint set' for jitted code
* Updated after review
* Enlarge buffer for line program
Hyung-Kyu Choi [Tue, 6 Sep 2016 17:13:31 +0000 (02:13 +0900)]
Move genIncRegBy() to codegenclassic.h (#7063)
We don't have to keep genIncRegBy() in codegen.h anymore,
since genIncRegBy() is defined in codegenlegacy.cpp and only used in LEGACY_BACKEND code.
Signed-off-by: Hyung-Kyu Choi <hk0110.choi@samsung.com>
Hanjoung Lee [Tue, 6 Sep 2016 16:42:03 +0000 (01:42 +0900)]
[ARM] Remove ARM_HAZARD_AVOIDANCE (#7019)
remove `ARM_HAZARD_AVOIDANCE` as it is a rare case.
For #7003
Peter Kukol [Tue, 6 Sep 2016 16:21:13 +0000 (10:21 -0600)]
Enable FEATURE_JIT_METHOD_PERF always (even when FEATURE_CORECLR is on). (#7028)
Joseph Tremoulet [Fri, 2 Sep 2016 19:22:01 +0000 (15:22 -0400)]
Fix GTF_ flag collision
Define flag `GTF_IS_IN_CSE` (which is used only to communicate in a call to
`gtNodeHasSideEffects` that the caller wants to disregard the side-effect
of running a .cctor) as `GTF_BOOLEAN` rather than `GTF_MAKE_CSE` -- as the
comment at the definition of `GTF_IS_IN_CSE` indicates, the only
requirement on its value is that it not conflict with one of the
side-effect flags, but `gtNodeHasSideEffects` does in fact check for
`GTF_MAKE_CSE`, so the current value conflicts.
Also update the code in `gtNodeHasSideEffects` that is dictated by
`GTF_IS_IN_CSE` to check just for that bit, instead of checking that the
caller passed in exactly `GTF_PERSISTENT_SIDE_EFFECTS_IN_CSE` -- no need
to entangle conceptually independent parameters.
Fixes #7042.
Sung-Jae Lee [Tue, 6 Sep 2016 15:45:33 +0000 (00:45 +0900)]
Add proper support for custom installed llvm tool chain. (#7050)
* Add LLVM_HOME enviroment variable support
Hyeongseok Oh [Tue, 6 Sep 2016 15:43:39 +0000 (00:43 +0900)]
fix helpers' flag: ULDiv, ULMod can throw Div-by-zero exception (#7052)