Jay Krell [Sat, 31 Aug 2019 17:35:14 +0000 (10:35 -0700)]
Rename parameter back.
Commit migrated from https://github.com/mono/mono/commit/
24180e9106d1940795363ae70634ebe1d1446fb3
Jay Krell [Sat, 31 Aug 2019 17:34:32 +0000 (10:34 -0700)]
PR: And then rename the icall back.
Commit migrated from https://github.com/mono/mono/commit/
5d9ca2064652b75f221e633122ec34f80fb50676
Jay Krell [Sat, 31 Aug 2019 17:13:47 +0000 (10:13 -0700)]
PR: Rename function; cleanup slightly.
Commit migrated from https://github.com/mono/mono/commit/
ea0db154b44be82f48cd13c85619f0e82295d511
Jay Krell [Sat, 31 Aug 2019 12:16:07 +0000 (05:16 -0700)]
[Coop] Monitor.
Commit migrated from https://github.com/mono/mono/commit/
91445e2286023292b219f3733bc629ed65f06e96
Jay Krell [Sat, 31 Aug 2019 07:40:33 +0000 (00:40 -0700)]
Follow convention -- "static inline" is only for .h files, "static" is for .c. (mono/mono#15502)
inlining is up to the compiler to decide.
It is used in .h files for linkage/dedup, not to affect codegen.
Commit migrated from https://github.com/mono/mono/commit/
a768b1533051aa37fc34af438e9af5bab48d515d
Jay Krell [Sat, 31 Aug 2019 07:38:12 +0000 (00:38 -0700)]
[coop] InternalGetHashCode (mono/mono#16583)
Commit migrated from https://github.com/mono/mono/commit/
cba94f28bed566e7a976842d016a7b2d4ef3f91a
Aleksey Kliger (λgeek) [Sat, 31 Aug 2019 06:06:25 +0000 (02:06 -0400)]
[runtime] Make eglib tests run on CI (mono/mono#16187)
[runtime] Make eglib tests run on CI
Update mono/eglib SUBDIRS to build tests after eglib is built.
Add a check-local automake rule.
Add CI steps to run the eglib tests.
Commit migrated from https://github.com/mono/mono/commit/
8bc27f6665fdd8464f3fdad15a974e5f38f9b7f2
Jay Krell [Fri, 30 Aug 2019 21:45:55 +0000 (14:45 -0700)]
Cleanup x86_patch and amd64_patch. (mono/mono#16474)
* Cleanup x86_patch and amd64_patch.
amd64_patch has two incorrect looking probably unused cases.
g_assert(0) and fix them.
call and jmp indirect (one case)
move reg RIP relative
x86_patch is large-ish inline macro.
Presumably for sharing with amd64, not inlining.
Restructure as an inline function, called once.
Try to make the code a little clearer also, with same functionality and efficiency
(modulo the optimized perf of "if i == 4" vs. "if i")
i.e. speak of instruction_size and offset_size, not pos and boolean size.
Add to x86_patch the ability to patch jmp/call indirect,
as amd64_patch has, but which was incorrect.
This would seem useful for a later change (using "thunks"
to patch nearby data instead of patching aligned instructions,
indirect call/jmp vs. direct).
amd64_patch largely delegates to x86_patch.
One case asserts and delegates.
Move the assert to the x86 side, where an optimizing
compiler should know to remove it for x86 anyway.
Change some assert to g_assert.
Remove some unnecessary casts.
Fix an assert (https://github.com/mono/mono/pull/16471).
* Change g_assert back to assert except where making larger changes
and less expansion.
Commit migrated from https://github.com/mono/mono/commit/
f735571f27738e3354f6d92650a0517d0482039d
Bernhard Urban [Fri, 30 Aug 2019 20:46:10 +0000 (22:46 +0200)]
[mini] publish global patches after JitInfo has been added (mono/mono#16589)
[mini] publish global patches after JitInfo has been added
Fixes https://github.com/mono/mono/issues/14080
Consider the following example:
```csharp
static void CommonCallTarget () { }
static void TailA () {
tailcall CommonCallTarget ();
}
static void TailB () {
tailcall CommonCallTarget ();
}
```
Since `TailA` and `TailB` are tailcalling into `CommonCallTarget`, the resolution at patch-time is a bit tricky, that is, since it's a jump-like instruction the patching machinery won't know where it was called from. That's why we maintain a global hashtable `jump_target_hash` where each jump-site is signed up to be patched. At patch-time we know the target method (in the example `CommonCallTarget`), but since we don't know where we are coming from, we will just apply all patches for that target.
This works since ages, so why did it crash on arm64 sometimes?
When the patching happens, we check if the displacement between jump-site and target fits into it (24bit). If not, which happens not very often, we have to allocate a _thunk_:
https://github.com/mono/mono/blob/mono/mono@
36296ce291f8a7b19de3eccb7a32c7e4ed9df8f2/mono/mini/mini-arm64.c#L928-L942
So instead of jumping to the target directly, we'll branch to the thunk. This is a little trampoline that loads the full address of the target and then finally branches to it. This one will live close-by the jump-site, because during compilation we will reserve specifically for that scenario some space after the generated code. For this, however, we need the JitInfo of the jump-site. And that's where the origin of the race is. Let's say:
* Thread A compiles `TailA`, and then jumps into it. Thus one patch point is in the `jump_target_hash`.
* Now Thread B compiles `TailB`, registers the patch point but has _not_ yet registered its JitInfo.
* Then Thread A continues, does the tailcall into `CommonCallTarget`, enters the patching machinery, which sees two patches. Now assume when applying the patch for `TailB` the displacement is too large, thus it tries to allocate a thunk but can't find the relevant JitInfo for it that it needs to emit the thunk. So it crashes as reported in https://github.com/mono/mono/issues/14080
As far as I can tell this only affects ARM64, ARM and PPC.
<!--
Thank you for your Pull Request!
If you are new to contributing to Mono, please try to do your best at conforming to our coding guidelines http://www.mono-project.com/community/contributing/coding-guidelines/ but don't worry if you get something wrong. One of the project members will help you to get things landed.
Does your pull request fix any of the existing issues? Please use the following format: Fixes #issue-number
-->
Commit migrated from https://github.com/mono/mono/commit/
06e63b345f0be60008597784d1cc8b3251bf60fa
Steve Pfister [Fri, 30 Aug 2019 20:27:14 +0000 (16:27 -0400)]
Turned the compiler server off on Windows because Named Pipes aren't fully working (mono/mono#16552)
* Turned the compiler server off on Windows because Named Pipes aren't fully working
* Added win32 to the CS is turned off message
Commit migrated from https://github.com/mono/mono/commit/
dbce6ac447dcb9443f50da9fe50caed50adc1d06
Bernhard Urban [Fri, 30 Aug 2019 20:26:12 +0000 (22:26 +0200)]
Revert "[android] add hardware specific workaround for Nexus9 in armv7 mode" (mono/mono#14691)
Revert "[android] add hardware specific workaround for Nexus9 in armv7 mode"
This reverts commit mono/mono@
ee90fc601e762dc31eeb90fc61b6305246ad5275 (also see https://github.com/mono/mono/pull/4878 ).
The device was released November 2014. I think it's fair to assume it isn't widely used anymore.
Commit migrated from https://github.com/mono/mono/commit/
5eb67ac88a2b26f28091d4f4876720ba3dbd8674
Zoltan Varga [Fri, 30 Aug 2019 18:08:41 +0000 (14:08 -0400)]
[netcore] Fix make patch-local-dotnet if .dotnet has more than one netcore install. (mono/mono#16590)
Commit migrated from https://github.com/mono/mono/commit/
4997d832916d092ee5a69bd28d0ee354a49c6e35
Jay Krell [Fri, 30 Aug 2019 17:30:58 +0000 (10:30 -0700)]
[coop] mono_runtime_class_init_full (mono/mono#16580)
Commit migrated from https://github.com/mono/mono/commit/
86c45ff645e59c997341ac44e95d03d4491c50eb
Nikolay Sivov [Fri, 30 Aug 2019 16:29:54 +0000 (19:29 +0300)]
[cominterop] Respect ComDefaultInterfaceAttribute when constructing ccw. (mono/mono#16584)
Commit migrated from https://github.com/mono/mono/commit/
c8707c6e84ed5978fb21be3c2dd4fe678b904c9c
Jay Krell [Fri, 30 Aug 2019 14:53:40 +0000 (07:53 -0700)]
[Coop] ves_icall_RuntimeFieldInfo_GetRawConstantValue. (mono/mono#16582)
Commit migrated from https://github.com/mono/mono/commit/
9db00e74950a5b9572dbcfb3c6403503ae5b236d
Bernhard Urban [Fri, 30 Aug 2019 13:16:03 +0000 (15:16 +0200)]
[configure] remove autodection of llvm-config in $PATH (mono/mono#16569)
[configure] remove autodection of llvm-config in $PATH
Instead `./autogen.sh --with-llvm=<path>` should be used if a build from `external/llvm` is not preferred.
The old behavior was confusing when a `llvm-config` is on `$PATH`, which can be the case when you have LLVM packages installed on a Linux distribution. It picked that up when `./autogen.sh --enable-llvm` was used.
/cc @brianrob
Commit migrated from https://github.com/mono/mono/commit/
07045604cf8e436359d04f18eff05ae00668e752
Jay Krell [Fri, 30 Aug 2019 08:27:19 +0000 (01:27 -0700)]
[interp] Compress interpreter opcode names in usual way, offsets vs. pointers. (mono/mono#16350)
* [interp] Compress the interpreter opcode names in the usual way, by changing
from an array of pointers to an array of 16 bit offsets into one large "string".
* PR: Make mono_interp_opname a function.
Commit migrated from https://github.com/mono/mono/commit/
36296ce291f8a7b19de3eccb7a32c7e4ed9df8f2
Jay Krell [Fri, 30 Aug 2019 03:51:15 +0000 (20:51 -0700)]
[coop] Convert System_Runtime_Remoting_Messaging_AsyncResult, etc. (mono/mono#16566)
* [coop] Convert System_Runtime_Remoting_Messaging_AsyncResult_Invoke, etc.
* Remove mono_load_remote_field_checked from this PR.
* Pin ac->msg earlier (should not matter).
* Add missing barrier.
* Pin slightly earlier -- still should not matter.
Commit migrated from https://github.com/mono/mono/commit/
3db78e829c0909bf9ca459e3a3e1033365386569
Jay Krell [Wed, 28 Aug 2019 00:20:02 +0000 (17:20 -0700)]
[Coop] mono_runtime_class_init_full mostly single point of exit.
Commit migrated from https://github.com/mono/mono/commit/
cbb9c645eb4d4d0f03da3dab3c99a4bcc9919fbe
Zoltan Varga [Fri, 30 Aug 2019 00:46:30 +0000 (20:46 -0400)]
Bump ikvm to pick up mono/mono@
caa8e7f54279a47422626005d228447f5d3670d5. (mono/mono#16567)
* Bump ikvm to pick up mono/mono@
caa8e7f54279a47422626005d228447f5d3670d5.
* [monodis] Add a missing cattr parent case.
Commit migrated from https://github.com/mono/mono/commit/
796a5751315dbf95bc7dedea684840e4cce84861
Egor Bogatov [Thu, 29 Aug 2019 23:01:15 +0000 (02:01 +0300)]
Make Math.Pow, Min and Max for doubles intrinsics (mono/mono#16561)
* Add new Math intrinsics
* call EB.setTargetOptions only for fast-math
* Address feedback
* rename float opcodes
* Rename opcodes
* Update CoreFX.issues.rsp
Commit migrated from https://github.com/mono/mono/commit/
c3ccb6e0272a6b000a8adfb81c02c240adffc649
Zoltan Varga [Thu, 29 Aug 2019 23:00:59 +0000 (19:00 -0400)]
[simd] Implement .ctor(T[]) and CopyTo () as intrinsics. (mono/mono#16558)
Commit migrated from https://github.com/mono/mono/commit/
210b3243471a61024d54b16d4092e90134ff098e
Jay Krell [Thu, 29 Aug 2019 20:16:14 +0000 (13:16 -0700)]
Cleanup trampoline size checks. (mono/mono#15624)
Cleanup trampoline size checks.
- named `int` everywhere
- `assert` everywhere
- assert `<=` everywhere instead of `<`
- `const` almost everywhere
Names not consistent, following existing code (`size`, `buf_len`, `kMaxCode`.)
Only touching x86/amd64, granted.
Could be better -- make them two pass and avoid the presizing.
Commit migrated from https://github.com/mono/mono/commit/
5390b796dd9c16eaded61c406a4cea9c66ea8905
Ryan Lucia [Thu, 29 Aug 2019 16:35:15 +0000 (12:35 -0400)]
[netcore] Implement AssemblyLoadContext.LoadUnmanagedDllFromPath (mono/mono#16525)
* [netcore] Implement AssemblyLoadContext.LoadUnmanagedDllFromPath
* Fixes
Commit migrated from https://github.com/mono/mono/commit/
3b57849541b35a8f6cced2957b9a6fdc0680136f
Nikolay Sivov [Thu, 29 Aug 2019 12:19:09 +0000 (15:19 +0300)]
[cominterop] Implement native-to-managed safearray marshalling. (mono/mono#16483)
* [cominterop] Implement native-to-managed safearray marshalling.
* [tests] Fix index variable type.
In attempt ot fix MSVC build failure (int* vs LONG* on Windows).
* [cominterop] Use proper statement block for newly added switch entry.
* [marshal] Conditionally compile in mono_cominterop_emit_marshal_safearray().
Hopefully should fix remaining build failures.
* [cominterop] Don't free user array argument
Commit migrated from https://github.com/mono/mono/commit/
421bd79f06a92372e0259be5349e4236edf0e7bb
Jay Krell [Thu, 29 Aug 2019 09:00:44 +0000 (02:00 -0700)]
[Coop] mono_load_remote_field_checked single point of exit. (mono/mono#16526)
Extracted from earlier version of mono/mono#16504.
Baby steps.
Commit migrated from https://github.com/mono/mono/commit/
52c1d4acfc751dd83f1964648cd91288988d7eec
Egor Bogatov [Thu, 29 Aug 2019 08:32:43 +0000 (11:32 +0300)]
LLVM: Add more Math intrinsics, introduce `--cpu` flag (mono/mono#16309)
* Add Math Intrinsics, introduce --cpu flag
* cpu -> mcpu
Commit migrated from https://github.com/mono/mono/commit/
2e9c5df855566402a314864580ec7dd908be2112
Jay Krell [Thu, 29 Aug 2019 08:23:20 +0000 (01:23 -0700)]
[coop] Cleanup mono_reflection_get_custom_attrs_blob_checked. (mono/mono#16535)
- single point of exit
- local variable cheaper than strcmp
Commit migrated from https://github.com/mono/mono/commit/
2f2d9541e593ed6976b32bde80a5f47b25ff9148
Filip Navara [Thu, 29 Aug 2019 00:35:07 +0000 (02:35 +0200)]
[netcore] Add Monitor.LockContentionCount icall (mono/mono#16538)
* Add Monitor.LockContentionCount icall
* Add NOHANDLES around the new icall
Commit migrated from https://github.com/mono/mono/commit/
e32fb11dbea1743b495241dc648065fa1938aa54
Jay Krell [Wed, 28 Aug 2019 22:10:40 +0000 (15:10 -0700)]
Read only function pointers (MonoThreadInfoRuntimeCallbacks). (mono/mono#16545)
Read only memory is preferred, for efficiency and security.
Even if it requires relocation first.
If the runtime is loaded in more than one process, physical pages can be shared (at least on Windows, where the kernel does the relocation).
If there is memory corruption, writes to here will sigsegv/accessviolate.
The older code also copied the pointers. For lifetime purposes that could be useful, but generally is not. i.e. they live with the code they point to.
Writable function pointers are also particularly targets for attackers.
There's been tons of defenses put around them through the years.
Read only ones are not vulnerable -- unless the attacker escalates and calls mprotect first.
When there is a JIT, arguably there is just tons of attack surface.
But there is also AOT.
Structs of function pointers are usually read only in the wider world.
Commit migrated from https://github.com/mono/mono/commit/
7ca965e09c35b1f9a2a7a56f68f0be997f841754
Zoltan Varga [Wed, 28 Aug 2019 20:16:18 +0000 (16:16 -0400)]
Merge pull request mono/mono#16323 from EgorBo/ffast-math
Fast Math mode
Commit migrated from https://github.com/mono/mono/commit/
c4695b8b4440e0f80c4768379fa24ba78ebff9d2
Jay Krell [Wed, 28 Aug 2019 19:33:48 +0000 (12:33 -0700)]
[interp] Fix remaining MonoError reuse. (mono/mono#16533)
* [interp] Fix remaining MonoError reuse.
This finishes off:
https://github.com/mono/mono/commit/mono/mono@
38aebeae38704df53ef8f2991ff63e293a3ac4de
which is *ok*, but leaves alone the other approach:
https://github.com/mono/mono/commit/mono/mono@
4d9e6ae20e536394c9ff65c61238045db294c08e
https://github.com/mono/mono/commit/mono/mono@
b654aa1ca858666539a677e96d00f594fd81c630
https://github.com/mono/mono/commit/mono/mono@
fd4956fbf277ec56a13137e01b48efb98a5556f5
Several other options are possible.
* self and peer PR: No need to cleanup successful error.
Commit migrated from https://github.com/mono/mono/commit/
056d5ca545d8534fa4b9c8b9268913775607e25a
Jay Krell [Wed, 28 Aug 2019 17:35:10 +0000 (10:35 -0700)]
Portably pass a full pointer and not just an integer to terminate vararg pointer lists. (mono/mono#15629)
i.e. because of define NULL 0.
Though I think at least Visual C++ special cases this and passes a full pointer to varargs.
Commit migrated from https://github.com/mono/mono/commit/
ff73c84482f99036f04c883316246d43d17f44d4
Jay Krell [Wed, 28 Aug 2019 16:59:11 +0000 (09:59 -0700)]
[interp] Possible code size savings via goto instead of macros. (mono/mono#16531)
[interp] Possible code size savings via goto instead of macros.
Possibly the compilers do this already.
Commit migrated from https://github.com/mono/mono/commit/
a0ba2cbce52dd312fb34a39d366dde5b4a878fe3
Egor Bogatov [Wed, 28 Aug 2019 16:01:34 +0000 (19:01 +0300)]
Merge remote-tracking branch 'origin/master' into ffast-math
Commit migrated from https://github.com/mono/mono/commit/
ffc3081c3b9576c922286c79f51fe3fb3cd49f50
imhameed [Wed, 28 Aug 2019 15:23:43 +0000 (08:23 -0700)]
Merge pull request mono/mono#16242 from lateralusX/lateralusX/windows-fix-local-valuetype-copy-llvm
OP_LLVM_OUTARG_VT doesn't always create temporary vt needed by some ABI's.
Commit migrated from https://github.com/mono/mono/commit/
bd3cc61301cc12d167b284316e9d99af92f2130b
Bernhard Urban [Wed, 28 Aug 2019 15:19:39 +0000 (17:19 +0200)]
[mini] do not re-AOT whole BCL for fullaotcheck targets (mono/mono#16536)
[mini] do not re-AOT whole BCL for fullaotcheck targets
And enable `llvmfullaotcheck` target on CI.
Fixes https://github.com/mono/mono/issues/15999
On my armv7/linux machine with a `--with-runtime-preset=fullaot_llvm` configured build, before this PR:
```console
$ time make -C mono/mini llvmfullaotcheck -j4
[...]
real 9m42.362s
user 23m2.000s
sys 0m18.480s
```
After:
```console
$ time make -C mono/mini llvmfullaotcheck -j4
[...]
real 3m1.984s
user 4m51.648s
sys 0m8.248s
```
Commit migrated from https://github.com/mono/mono/commit/
71f0f0ff827b03252e8967aeb6152a14cc3198ec
Alexander Köplinger [Wed, 28 Aug 2019 14:17:31 +0000 (16:17 +0200)]
[netcore] Add missing 'prepare' dependency to run-tests-corefx
Commit migrated from https://github.com/mono/mono/commit/
2152f0a16d68021fa28a997806992ad59fd0c7e7
Zoltan Varga [Wed, 28 Aug 2019 13:46:09 +0000 (09:46 -0400)]
Merge pull request mono/mono#16517 from vargaz/wasm-xunit4
[wasm] Enable xunit tests for corlib/System.Core.
Commit migrated from https://github.com/mono/mono/commit/
f562c7ef6a23e9f713a53b446e937eedc63728b6
Jay Krell [Wed, 28 Aug 2019 13:42:44 +0000 (06:42 -0700)]
Fix assert in x86_patch regarding second byte of Jcc instructions being 0x80-0x8F vs. 0x70-0x8F (6byte conditional branch with 32bit offset). (mono/mono#16471)
Commit migrated from https://github.com/mono/mono/commit/
c80a1eca636c82014f8fbb9873e5391cddffa9a9
Aleksey Kliger (λgeek) [Wed, 28 Aug 2019 12:34:30 +0000 (08:34 -0400)]
[merp] Initialize hang_watchdog_path earlier; be lax if it's missing (mono/mono#16298)
* [merp] Initialize hang_watchdog_path earlier; be lax if it's missing
Initialize hang_watchdog_path earlier - not just when
ves_icall_Mono_Runtime_EnableCrashReportingLog is called. That function is
only expected to be called by apps that explicitly want progress reports from
MERP when it's collecting a crash report. But we want the mono-hang-watchdog
to execute for all mono crashes (when crash reporting isn't ifdef'd out).
Also if we try to exec mono-hang-watchdog and fail for some reason, print a
nice message and then exit. We used to call g_assert_not_reached which would
kick off another MERP of the _forked child process_ which would then get
confused because it really can't run in the forked process at all.
Commit migrated from https://github.com/mono/mono/commit/
fc21168065af270d1e500fa6be618f8563d0c525
Jay Krell [Wed, 28 Aug 2019 10:51:23 +0000 (03:51 -0700)]
[interp] Outline box. (mono/mono#16456)
With additional reasonable changes this should conserve stack.
A combined PR that demonstrates actual savings probably should/will be presented.
Commit migrated from https://github.com/mono/mono/commit/
2ca04a9743ceab99714377180ba3b2b6d589ba7f
Jay Krell [Wed, 28 Aug 2019 10:11:56 +0000 (03:11 -0700)]
[interp] Minor cleanup of `stackval_from_data`. Compiler can do it. (mono/mono#16534)
I missed this earlier cleaning stackval_to_data.
Commit migrated from https://github.com/mono/mono/commit/
56b83f489b42e076bd7259cf42b4c684e63c7785
Jay Krell [Wed, 28 Aug 2019 10:10:30 +0000 (03:10 -0700)]
[interp] Asserts pessimize clang. Removing saves 16 bytes on Linux/amd64/clang. (mono/mono#16530)
Paired with https://github.com/mono/mono/pull/16456, saves 16 bytes on Linux/amd64/clang.
No change with gcc.
Contributes to https://github.com/mono/mono/issues/16172.
Commit migrated from https://github.com/mono/mono/commit/
c870df0aa7a8111743bab8bbe6b9f3f6fa8ed870
Jay Krell [Wed, 28 Aug 2019 08:56:08 +0000 (01:56 -0700)]
Remove dead Boehm code. (mono/mono#15224)
* Remove dead Boehm code.
Unity looked similarly dead.
The key points making it dead were:
create_allocator:
g_assert_not_reached() covering everything but local variables
mono_gc_get_managed_allocator:
always returning NULL, killing a bunch of code
after it
mono_gc_is_critical_method:
looking at data that was never filled in
This should have no impact, not even on Boehm.
Boehm should continue working just the same.
It is nice because it is the few/only uses of CEE_MONO_TLS.
Commit migrated from https://github.com/mono/mono/commit/
c534511003bdd84ef08dd8a7cf725d4d0f063f5e
Jay Krell [Wed, 28 Aug 2019 08:54:00 +0000 (01:54 -0700)]
[mini] Cleanup MONO_MMAP_32BIT. (mono/mono#16469)
- It _presumably_ means nothing on 32bit ARM and 32bit MIPS (was there mono/mips64?).
- Leaving only PowerPC32 and PowerPC64 (in this vicinity).
- It presumably means nothing on PowerPC32.
- Leaving PowerPC64 as the only user in this vicinity -- is it needed?
- Notice this is varyingly absent on x86, arm64, amd64, s390x.
Subject of another PR maybe.
Perhaps this was to enable the debugger to be 32bit against a 64bit target
on a biarch system, so it could do a cross-process manipulation of these pages?
That makes some sense on Windows at least.
(32bit debugger cannot debug 64bit target as of Vista, but that is using the Win32
"hard" debugger API. Mono could do otherwise).
There remains amd64 mono-codeman.c define ARCH_MAP_FLAGS MONO_MMAP_32BIT left alone.
Commit migrated from https://github.com/mono/mono/commit/
0375acd7c8b1634a44477f2a561a10cd4d01858a
Zoltan Varga [Wed, 28 Aug 2019 07:24:01 +0000 (03:24 -0400)]
Merge pull request mono/mono#16520 from vargaz/fix-llvm-2
[llvm] Fix ctlz intrinsics, they have 2 arguments.
Commit migrated from https://github.com/mono/mono/commit/
162849f97202e8a24355fb77b01657b891436383
Bernhard Urban [Tue, 27 Aug 2019 19:20:17 +0000 (21:20 +0200)]
fix cast and signature for mono_rconv_u4
Commit migrated from https://github.com/mono/mono/commit/
d6fa5ff9c6c482582149c93613468a0e0432d2af
Bernhard Urban [Tue, 27 Aug 2019 16:38:57 +0000 (09:38 -0700)]
[amd64] use opcode emulation for fconv/rconv
Commit migrated from https://github.com/mono/mono/commit/
460bfa9beb5159156df2256766d738debeffe706
Bernhard Urban [Tue, 27 Aug 2019 16:37:46 +0000 (09:37 -0700)]
[llvmaotcheck] require sse3 being available for regression tests
Commit migrated from https://github.com/mono/mono/commit/
8e6b967391533fbbe555e8aa83a5fdf501f81c36
Bernhard Urban [Tue, 27 Aug 2019 12:47:54 +0000 (08:47 -0400)]
[arm] use opcode emulation for fconv/rconv
Commit migrated from https://github.com/mono/mono/commit/
5b20322a84bac20433103ac06a59e654cf581485
Bernhard Urban [Tue, 27 Aug 2019 12:46:06 +0000 (08:46 -0400)]
[arm] account only 4 bytes on stack for single precision arguments
Fixes `test_0_arm64_small_stack_args` in gshared.cs when
* `test_0_arm64_small_stack_args` is compiled with LLVM
* `Foo3.Floats` is compiled with Mini
Commit migrated from https://github.com/mono/mono/commit/
ba824cb9213670069b85ab397a7163a50e911ff6
Bernhard Urban [Mon, 26 Aug 2019 20:00:06 +0000 (16:00 -0400)]
[arm64] use opcode emulation for fconv/rconv
LLVM eagerly optimizes constants such as NaN in a way that does not align with .NET behaviour. In the future we can change it so that we only do opcode emulation when LLVM is used.
Fixes https://github.com/mono/mono/issues/16411
Commit migrated from https://github.com/mono/mono/commit/
475defca4d4d24dc44c660244b645df5f33e20ac
Jay Krell [Mon, 26 Aug 2019 06:05:18 +0000 (23:05 -0700)]
Cleanup arm64 SeqPointInfo.
Commit migrated from https://github.com/mono/mono/commit/
645c01066c08adef6d302a25d1ed9db6047e3042
Jay Krell [Tue, 27 Aug 2019 21:39:18 +0000 (14:39 -0700)]
Inline mono_string_length_internal and mono_string_chars_internal (mono/mono#16374)
* Inline mono_string_length_internal and mono_string_chars_internal for everyone.
* Add MONO_REQ_GC_UNSAFE_MODE.
Commit migrated from https://github.com/mono/mono/commit/
f516e79dd981e174ebf04fe81b76cdbec25f6b20
Filip Navara [Tue, 27 Aug 2019 20:55:37 +0000 (22:55 +0200)]
Update version information in System.Private.CoreLib to mimic CoreCLR and make BenchmarkDotNet happy (mono/mono#16515)
BenchmarkDotNet recently updated the algorithms for .NET version detection (https://github.com/dotnet/BenchmarkDotNet/pull/1230) which made it incompatible with the version presented by Mono.
This updates the value of [ProductName](https://github.com/dotnet/coreclr/blob/mono/mono@
1f428dbc5554191dadd27511cd5cd181e4a60efb/Directory.Build.targets#L7) and [Version](https://github.com/dotnet/coreclr/blob/mono/mono@
c9007e73286456434ba12744e778e87a89fc1d90/eng/Versions.props#L5) to match what CoreCLR is currently reporting and makes the [BDN detection code](https://github.com/dotnet/BenchmarkDotNet/blob/mono/mono@
b6d0e0f1d80ed60aa157149a0dfb704a2da4ca41/src/BenchmarkDotNet/Environments/Runtimes/CoreRuntime.cs#L73-L81) happy.
Commit migrated from https://github.com/mono/mono/commit/
f4145ce10815b2f4a1d585dbdf462aa17015aed1
Ryan Lucia [Tue, 27 Aug 2019 20:53:37 +0000 (16:53 -0400)]
[netcore] Change exception thrown for SetTypedReference (mono/mono#16518)
This brings us in line with CoreCLR behavior.
Commit migrated from https://github.com/mono/mono/commit/
f9f5fff40c2012dc9631b7128d72c945620ce4fb
Zoltan Varga [Tue, 27 Aug 2019 20:34:27 +0000 (16:34 -0400)]
[llvm] Fix ctlz intrinsics, they have 2 arguments.
Commit migrated from https://github.com/mono/mono/commit/
340289d4aed8b8fe59379991d7b1320b419a2d36
Zoltan Varga [Tue, 27 Aug 2019 20:31:18 +0000 (16:31 -0400)]
[llvm] Fix the mixing of LLVM modules in the llvm JIT, use ctx->lmodule instead of module->lmodule.
Commit migrated from https://github.com/mono/mono/commit/
ab1e9be35e78e6a99f5e97cc6e1df514ddbdb8ab
Zoltan Varga [Tue, 27 Aug 2019 20:30:16 +0000 (16:30 -0400)]
Fix the netcore build. (mono/mono#16519)
Commit migrated from https://github.com/mono/mono/commit/
44d33673a5d390e1f691967eb776c608d60b7a64
Zoltan Varga [Tue, 27 Aug 2019 19:22:30 +0000 (15:22 -0400)]
Merge pull request mono/mono#16487 from vargaz/cattr-data-handles
[runtime] Convert mono_reflection_create_custom_attr_data_args () to use handles.
Commit migrated from https://github.com/mono/mono/commit/
3108d4305174a015484aa0f77226e98ea13cb948
Egor Bogatov [Tue, 27 Aug 2019 17:21:51 +0000 (20:21 +0300)]
use gunichar2
Commit migrated from https://github.com/mono/mono/commit/
e9dff75d2fb049f11fdcf595b68de1f3061bdeb7
Alexander Köplinger [Mon, 26 Aug 2019 23:50:15 +0000 (01:50 +0200)]
Remove AppContext.cs that was moved to shared
Commit migrated from https://github.com/mono/mono/commit/
f914d900f6fc639b25eca68432133fedd4f6297b
Vlad Brezae [Tue, 27 Aug 2019 17:27:22 +0000 (20:27 +0300)]
[interp] Remove unnecessary recomputation (mono/mono#16509)
Seems to have been added by mistake in https://github.com/mono/mono/pull/16317, which was meant just to remove the locals variable from InterpFrame.
Commit migrated from https://github.com/mono/mono/commit/
ccc4bf853618649602691c6a10af59f80cd18466
Jay Krell [Tue, 27 Aug 2019 13:31:10 +0000 (06:31 -0700)]
[interp] Error handling refactor (optimize non-exception path) (helps clang save stack) (mono/mono#16371)
While working on stack reduction, I believe I found that
the compiler could not tell how THROW_EX and/or NULL_CHECK terminate.
That is, essentially, the success paths preserve volatile registers.
They do not have to be spilled/filled.
This tweak should make it clearer to the compiler, optimized common paths, and not slow down uncommon paths.
This also cleans it up a little and reduces macro expansion, and then possibly should allow for reinlining what is now used only once.
Also add some const so it can be casted away less.
Believed to save 32 bytes of stack with Linux/amd64/clang, nothing for gcc.
And a nice cleanup.
Commit migrated from https://github.com/mono/mono/commit/
00c1ec421a3d47931bd9ad719faec62e60277837
Zoltan Varga [Tue, 27 Aug 2019 10:51:03 +0000 (06:51 -0400)]
Merge pull request mono/mono#16493 from EgorBo/impl-lzcnt
Implement System.Runtime.Intrinsics.X86.Lzcnt
Commit migrated from https://github.com/mono/mono/commit/
fa1288794b5a38fa717d0f8db459386c3d4b3ad3
Johan Lorensson [Tue, 27 Aug 2019 10:14:12 +0000 (12:14 +0200)]
Fix multiple SGEN_LOG to correctly scale elapsed time in usec. (mono/mono#16485)
Several SGEN_LOG entries logged usec based on result returend from TV_ELAPSED.
On SGEN's implementation of TV_ELAPSED, the result is not scaled to usec
but returned as 100ns ticks. This was not handled correctly by SGEN_LOG,
but other measures, like profile counters and SGEN binary protocol handles
it correctly.
Fix will adjust the SGEN_LOG calls currently presenting usec to correctly
scale the result. Fix also adds one additional logging measuring total
amount of time spend scanning thread data.
Commit migrated from https://github.com/mono/mono/commit/
70aba13173840818246b470e425a6d2585e7672d
Jay Krell [Tue, 27 Aug 2019 09:51:21 +0000 (02:51 -0700)]
[interp] Replace imethod with frame->imethod. (mono/mono#16459)
This conserves stack on Linux/amd64/gcc.
Commit migrated from https://github.com/mono/mono/commit/
bfe1b6182cb4e354f94d00d3206be4b83331db1d
Jay Krell [Tue, 27 Aug 2019 08:40:35 +0000 (01:40 -0700)]
CodeManager cleanup: (mono/mono#16468)
- Chunks do not need flags, only codemanagers do.
- Sort by size to minimize padding for alignment.
- Booleans only need one bit.
- Chunk is not spelled chunck.
- Suggest foo accessors be named foo() and set_foo(), not get_foo() and set_foo().
i.e. the more common "get" is implied.
- Allow free(0) w/o initialization.
Commit migrated from https://github.com/mono/mono/commit/
6cf8b6e73824545709cf6556726e0460bc7e2558
Jay Krell [Tue, 27 Aug 2019 08:36:59 +0000 (01:36 -0700)]
[monodis] Combine adjacent identical cases. (mono/mono#16501)
Commit migrated from https://github.com/mono/mono/commit/
9f9e5f7803b98494ca8773876c81e0cc582d679d
Zoltan Varga [Tue, 27 Aug 2019 07:57:44 +0000 (03:57 -0400)]
Add SETUP_ICALL_FUNCTION.
Commit migrated from https://github.com/mono/mono/commit/
c7d785f32996e9faa693f5e297d2ac47ee795acc
Jay Krell [Tue, 27 Aug 2019 07:46:35 +0000 (00:46 -0700)]
mono_attach_load_agent: Fix leaks, alloc, copy. (mono/mono#15950)
Extracted and slightly modified from https://github.com/mono/mono/pull/15868.
Commit migrated from https://github.com/mono/mono/commit/
7ca15e5a42ad03572e9209abb7e030ad712ee4b5
Jay Krell [Mon, 26 Aug 2019 23:04:46 +0000 (16:04 -0700)]
Add missing const, particularly on functions "like" memcpy. (mono/mono#16477)
But do not change the signatures of external functions.
Commit migrated from https://github.com/mono/mono/commit/
67c97d30b45751b5d87254d07c5343d92008dfb5
Zoltan Varga [Mon, 26 Aug 2019 21:31:50 +0000 (17:31 -0400)]
Merge pull request mono/mono#16451 from vargaz/gc-handle-pin
[runtime] Make the local gc handles pin the object they point to by scanning the handle area conservatively.
Commit migrated from https://github.com/mono/mono/commit/
4f8c0b9b46eb014ba7db36d824b3002c217a6c5b
Aleksey Kliger (λgeek) [Mon, 26 Aug 2019 21:13:45 +0000 (17:13 -0400)]
[threadpool-io] Throw if backend won't be able to register a FD (mono/mono#16396)
The poll+select i/o selector backend can't handle file descriptor ids greater
than FD_SETSIZE. This can happen if too many files are open and we want to
wait on it.
Previously, mono would fail in the i/o selector thread by which point it was
too late to do anything.
With this change we will fail eagerly on the thread that calls IOSelector.Add
by throwing a NotSupportedException.
Addresses https://github.com/mono/mono/issues/15931
Commit migrated from https://github.com/mono/mono/commit/
78edafd6e2d5ce886e2fd2ddc3cf8bbbffeeb860
Zoltan Varga [Mon, 26 Aug 2019 18:17:09 +0000 (14:17 -0400)]
Fix review comments.
Commit migrated from https://github.com/mono/mono/commit/
65a71a2912920768134442e04ea17c4fa7382d22
Egor Bogatov [Mon, 26 Aug 2019 18:08:53 +0000 (21:08 +0300)]
Implement Lzcnt hw intrinsics
Commit migrated from https://github.com/mono/mono/commit/
044d1519bdf40f37ec06128dd383c21157dc3c0e
Egor Bogatov [Mon, 26 Aug 2019 15:35:38 +0000 (18:35 +0300)]
[netcore] Implement missing ThreadPool methods (for metrics) (mono/mono#16076)
* Implement missing ThreadPool methods
* Address feedback
* ifdef for mono_threadpool_worker_get_completed_threads_count
* remove GetPendingUnmanagedWorkItemCount
Commit migrated from https://github.com/mono/mono/commit/
85265d884cf31f8779daba2d639eb795fbc9bd5e
Zoltan Varga [Mon, 26 Aug 2019 15:17:15 +0000 (11:17 -0400)]
[runtime] Convert mono_reflection_create_custom_attr_data_args () to use handles.
Commit migrated from https://github.com/mono/mono/commit/
a908b72037d0fd13f2a86208f35e833b62a96e55
Johan Lorensson [Mon, 26 Aug 2019 14:46:43 +0000 (16:46 +0200)]
Fix special case tramp assert in x86/amd64. (mono/mono#16476)
https://github.com/mono/mono/pull/16408 temporary fixed a tramp issue on
x86 (also exist on amd64). This commit fixes the underlying issue and re-enable
the use of noreturn outside of netcore.
The problem hit by the use of OP_NOT_REACHED on x86/amd64 is due to a specific
case when the address to patch happens to fall at the start of a different
managed method. Since the use of OP_NOT_REACHED can end a method
with the call instruction on x86/amd64 in combination with how patch location
(using the return address from stack) is resolved, the implementation in
mono_arch_patch_callsite didn't take this case into account, reading incorrect
patch data, triggering assert oh x86. So, based on timing we could end up with
the following code:
025715B3 call
02570D98
025715B8 push ebp
since the patch target is
025715B8, but also the start of a completely different
method, method_start == origin_code. The fix is to detect this case on x86/amd64
(done on each arch since the way origin_code is detected is arch specific)
and then use a method_start of NULL, that is already a supported scenario.
Commit migrated from https://github.com/mono/mono/commit/
f9be7a67d33d9bdec342401a8ed84885672e9b67
Alexander Köplinger [Mon, 26 Aug 2019 11:53:40 +0000 (13:53 +0200)]
[utils] Don't use MAP_32BIT on Apple platforms, fixes crash with XCode 11 beta6 (mono/mono#16441)
As of XCode 11 beta6 the MacOSX SDK defines the MAP_32BIT symbol:
```diff
--- /Applications/Xcode11-beta5.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.15.sdk/usr/include/sys/mman.h 2019-07-25 17:43:49.
000000000 -0400
+++ /Applications/Xcode11-beta6.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.15.sdk/usr/include/sys/mman.h 2019-08-06 21:03:07.
000000000 -0400
@@ -1,5 +1,5 @@
/*
- * Copyright (c) 2000-2002 Apple Computer, Inc. All rights reserved.
+ * Copyright (c) 2000-2019 Apple Computer, Inc. All rights reserved.
*
* @APPLE_OSREFERENCE_LICENSE_HEADER_START@
*
@@ -145,6 +145,10 @@
#define MAP_RESILIENT_CODESIGN 0x2000 /* no code-signing failures */
#define MAP_RESILIENT_MEDIA 0x4000 /* no backing-store failures */
+#if !defined(CONFIG_EMBEDDED)
+#define MAP_32BIT 0x8000 /* Return virtual addresses <4G only: Requires entitlement */
+#endif /* !defined(CONFIG_EMBEDDED) */
+
#endif /* (!_POSIX_C_SOURCE || _DARWIN_C_SOURCE) */
/*
```
This causes the mono_valloc() function to try to use the MAP_32BIT flag for mmap().
However as mentioned in the comment for the symbol in mman.h it seems to require a special entitlement
which isn't available/documented anywhere yet.
This in turn causes the mmap call to fail presumably because we're missing that entitlement.
Instead we now skip setting this flag on Apple platforms to make mmap() behave like it did before.
Commit migrated from https://github.com/mono/mono/commit/
2f2771fcfa22ac16a24a6a1ff2554440e5de29ca
Egor Bogatov [Mon, 26 Aug 2019 11:36:04 +0000 (14:36 +0300)]
Optimized LLVM IR (mono/mono#16448)
LLVM JIT: dump IR after optimizations
```csharp
[MethodImpl(MethodImplOptions.NoInlining)]
static int Test(int a, int b, int c) => a * b + a * c;
```
`MONO_VERBOSE_METHOD="Test"` now prints both unoptimized and optimized IR:
```llvm
*** Unoptimized LLVM IR for P:Test ***
; Function Attrs: noinline uwtable
define monocc i32 @"P:Test (int,int,int)"(i32 %arg_a, i32 %arg_b, i32 %arg_c) mono/mono#0 {
BB0:
br label %BB3
BB3: ; preds = %BB0
br label %BB2
BB2: ; preds = %BB3
%t22 = mul i32 %arg_a, %arg_b
%t25 = mul i32 %arg_a, %arg_c
%t27 = add i32 %t22, %t25
br label %BB1
BB1: ; preds = %BB2
ret i32 %t27
}
***
*** Optimized LLVM IR for P:Test ***
; Function Attrs: noinline uwtable
define monocc i32 @"P:Test (int,int,int)"(i32 %arg_a, i32 %arg_b, i32 %arg_c) mono/mono#0 {
BB0:
%t251 = add i32 %arg_b, %arg_c
%t27 = mul i32 %t251, %arg_a
ret i32 %t27
}
***
```
Commit migrated from https://github.com/mono/mono/commit/
6aaec2fe9b0c57dcd03967160d8c67bebcd91fae
Jay Krell [Mon, 26 Aug 2019 08:26:07 +0000 (01:26 -0700)]
[interp] Cleanup stackval_to_data and stackval_to_data_addr. (mono/mono#16458)
No functional change, except for debugging, look up a frame
for the prior value.
Commit migrated from https://github.com/mono/mono/commit/
6671316e35d9753539a0276c2c90c0300403b0ad
Bernhard Urban [Mon, 26 Aug 2019 07:46:12 +0000 (09:46 +0200)]
[android] remove text relocations in AOT/arm32 (mono/mono#16432)
[android] remove text relocations in AOT/arm32
Remove support for "big assemblies" for AOT on Android running on ARM 32bit. This will remove relocations in `.text` which are not allowed in the upcoming Android Q release anymore.
_If_ a customer should run into this problem, the suggested workarounds are:
* Switch to ARM64, or
* Do not use AOT compilation.
Fixes https://github.com/mono/mono/issues/16369
Commit migrated from https://github.com/mono/mono/commit/
b6ca76c48415393f09f09977327c0bcd99f10bf4
Ryan Lucia [Mon, 26 Aug 2019 03:13:39 +0000 (23:13 -0400)]
Detect additional invalid assembly names when parsing (mono/mono#16446)
Commit migrated from https://github.com/mono/mono/commit/
510f3c7d70d576cffc5f32edf8741ed841e6211d
Jay Krell [Sun, 25 Aug 2019 21:52:17 +0000 (14:52 -0700)]
Add MONO_LLVM_INTERNAL to mono_jit_compile_method. (mono/mono#16463)
Commit migrated from https://github.com/mono/mono/commit/
e95ef822a16406c2c75f973f22c6e47d0e4f4e7f
Jay Krell [Sun, 25 Aug 2019 17:11:02 +0000 (10:11 -0700)]
[interp] Outline mono_interp_box_vt. (mono/mono#16454)
Contributes to https://github.com/mono/mono/issues/16172.
Seen to reduce frame from 0xB8 to 0xA8 on Linux/amd64/gcc against master.
Possibly similar elsewhere.
(Possibly 8 added to a rounded 8.)
Add const, add braces, they might help compiler and/or programmer.
Commit migrated from https://github.com/mono/mono/commit/
2224d6b55d286e8ab1f2b222f4171f9667c27ed6
Zoltan Varga [Sat, 24 Aug 2019 15:32:37 +0000 (11:32 -0400)]
[runtime] Make the local gc handles pin the object they point to by scanning
the handle area conservatively.
Commit migrated from https://github.com/mono/mono/commit/
7d43ffda1a29315b20d2ab64d19db9d33e8faa4d
Jay Krell [Sat, 24 Aug 2019 07:56:40 +0000 (00:56 -0700)]
[interp] Outline STRMFLD_VT. (mono/mono#16355)
Without being entirely scientific about it, this case seems
to be one that uses more than typical locals and therefore
be on a sort of critical path to reduce frame size.
I have a change that reduces frame to 0x78 bytes (albeit
not yet working) and to achieve that, in this case I refetched
the locals after function calls.
This achieves similar but perhaps more elegantly.
If we rest at larger frame size then this case might not be
on the critical path.
https://github.com/mono/mono/issues/16172
Commit migrated from https://github.com/mono/mono/commit/
9d1eec6bf88656e8302441d0349dcc467a52b605
Jay Krell [Sat, 24 Aug 2019 07:56:00 +0000 (00:56 -0700)]
[windows gc] Trim one syscall per thread in garbage collection. (mono/mono#16003)
* Skip one syscall per thread in garbage collection.
Specifically do not query for guard pages.
There are better ways:
1. Visual C++ specific, __try / __except, get the guard page exception,
put back the guard page. This is a little slower (exception + two syscalls),
but exceedinly rare. __try itself is cheap on x86, free on the rest.
This was also not allowed on UWP. It will be. And is VirtualProtectFromApp.
2. Record the TEB and get from it the maximum commited page and just use MAX, trivial.
The comments are also not completely clear as to what is going on here, but they were close.
Augment them.
Commit migrated from https://github.com/mono/mono/commit/
99e9d8064272eefdba7c567f44d924ad1ec938ac
Jay Krell [Sat, 24 Aug 2019 07:52:30 +0000 (00:52 -0700)]
Remove dlmalloc from Win32. (mono/mono#15773)
Commit migrated from https://github.com/mono/mono/commit/
ab6a09741d1017800ecc2d1688562f68547d7834
Zoltan Varga [Fri, 23 Aug 2019 22:06:17 +0000 (18:06 -0400)]
Merge pull request mono/mono#16414 from vargaz/llvm-gc-poll-wrapper
[llvm] Add a mono wrapper method for the gc poll cold cconv wrapper so the EH code can unwind through it.
Commit migrated from https://github.com/mono/mono/commit/
ab9c58d76d49c750bf69ee4f6ba89c45c8a1df51
Ryan Lucia [Fri, 23 Aug 2019 19:58:10 +0000 (15:58 -0400)]
Return an empty string for assemblies loaded from stream (mono/mono#16345)
* Return an empty string for assemblies loaded from memory
* Blind attempt to fix wasm
Commit migrated from https://github.com/mono/mono/commit/
83cf007ae3e52a7c83d2a31bbc939aa08df67392
Ryan Lucia [Wed, 21 Aug 2019 22:35:24 +0000 (18:35 -0400)]
[netcore] Use requesting assembly's context for InternalLoad calls
Commit migrated from https://github.com/mono/mono/commit/
7615ba3e45654ce5b22125c818876c10b6e4fbf9
Ryan Lucia [Wed, 21 Aug 2019 20:40:08 +0000 (16:40 -0400)]
[netcore] Random loader cleanup
Commit migrated from https://github.com/mono/mono/commit/
0661bf0424f78f35571c229733d61b406f082e5b
Zoltan Varga [Fri, 23 Aug 2019 12:13:21 +0000 (08:13 -0400)]
Merge pull request mono/mono#16431 from kjpou1/wasm-maindir-cleanup
[wasm] Small cleanup of main wasm directory
Commit migrated from https://github.com/mono/mono/commit/
f0cbb79b7f17d30c6674f600f4fccef8a4f1cc87
Zoltan Varga [Fri, 23 Aug 2019 11:49:20 +0000 (07:49 -0400)]
[jit] Make mono_unwind_frame () ignore registers which are greater than NUM_DWARF_REGS.
These registers can be encountered in the unwind info for functions with a cold
cconv, like xmm0 on amd64. Since they are caller saved, they don't have to
be restored during EH.
Commit migrated from https://github.com/mono/mono/commit/
97d2d976156f5584d913844f2610683978f2a567
Zoltan Varga [Fri, 23 Aug 2019 10:40:09 +0000 (06:40 -0400)]
Merge pull request mono/mono#16429 from vargaz/llvm-jit-module
[llvm] Don't pass a module argument to the code in llvm-jit.cpp, we create a new LLVM module for each method.
Commit migrated from https://github.com/mono/mono/commit/
fdd0d7f027f2f45a8aacd304747a42e4ae7d9739
Zoltan Varga [Fri, 23 Aug 2019 10:28:40 +0000 (06:28 -0400)]
Fix the bitcode build.
Commit migrated from https://github.com/mono/mono/commit/
81f736e7745d7d94c282eb20605e970db74aeef1
Zoltan Varga [Fri, 23 Aug 2019 09:34:52 +0000 (05:34 -0400)]
[llvm] Add a cold wrapper and mono wrapper for gc poll for JITted code as well.
Commit migrated from https://github.com/mono/mono/commit/
4ed1c3335f2ddf88e1e402ad046bcb2d7cebc7d0