Summary: I think it would be better to require the alignment to be >= 1. It is currently confusing to allow both values.
Reviewers: courbet
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D77372
This will likely introduce catastrophic performance regressions on
older subtargets, but should be correct. A follow up change will
remove the old fp32-denormals subtarget features, and switch to using
the new denormal-fp-math/denormal-fp-math-f32 attributes. Frontends
should be making sure to add the denormal-fp-math-f32 attribute when
appropriate to avoid performance regressions.
Uploading output from `git format-patch` fails when version has
more than 2 dots, e.g. git version 2.24.1.windows.2 which is
currently recommended by e.g. GitExtensions or 2.24.1.rc on Linux.
Differential Revision: https://reviews.llvm.org/D72374
Add an option to llvm-dwarfdump to calculate the bytes within
the debug sections. Dump this numbers when using --statistics
option as well.
This is an initial patch (e.g. we should support other units,
since we only support 'bytes' now).
Differential Revision: https://reviews.llvm.org/D74205
Summary:
As noted in documentation, different repetition modes have different trade-offs:
> .. option:: -repetition-mode=[duplicate|loop]
>
> Specify the repetition mode. `duplicate` will create a large, straight line
> basic block with `num-repetitions` copies of the snippet. `loop` will wrap
> the snippet in a loop which will be run `num-repetitions` times. The `loop`
> mode tends to better hide the effects of the CPU frontend on architectures
> that cache decoded instructions, but consumes a register for counting
> iterations.
Indeed. Example:
>>! In D74156#1873657, @lebedev.ri wrote:
> At least for `CMOV`, i'm seeing wildly different results
> | | Latency | RThroughput |
> | duplicate | 1 | 0.8 |
> | loop | 2 | 0.6 |
> where latency=1 seems correct, and i'd expect the througput to be close to 1/2 (since there are two execution units).
This isn't great for analysis, at least for schedule model development.
As discussed in excruciating detail in
>>! In D74156#1924514, @gchatelet wrote:
>>>! In D74156#1920632, @lebedev.ri wrote:
>> ... did that explanation of the question i'm having made any sense?
>
> Thx for digging in the conversation !
> Ok it makes more sense now.
>
> I discussed it a bit with @courbet:
> - We want the analysis tool to stay simple so we'd rather not make it knowledgeable of the repetition mode.
> - We'd like to still be able to select either repetition mode to dig into special cases
>
> So we could add a third `min` repetition mode that would run both and take the minimum. It could be the default option.
> Would you have some time to look what it would take to add this third mode?
there appears to be an agreement that it is indeed sub-par,
and that we should provide an optional, measurement (not analysis!) -time
way to rectify the situation.
However, the solutions isn't entirely straight-forward.
We can just add an actual 'multiplexer' `MinSnippetRepetitor`, because
if we just concatenate snippets produced by `DuplicateSnippetRepetitor`
and `LoopSnippetRepetitor` and run+measure that, the measurement will
naturally be different from what we'd get by running+measuring
them separately and taking the min.
([[ https://www.wolframalpha.com/input/?i=%28x%2By%29%2F2+%21%3D+min%28x%2C+y%29 | `time(D+L)/2 != min(time(D), time(L))` ]])
Also, it seems best to me to have a single snippet instead of generating
a snippet per repetition mode, since the only difference here is that the
loop repetition mode reserves one register for loop counter.
As far as i can tell, we can either teach `BenchmarkRunner::runConfiguration()`
to produce a single report given multiple repetitors (as in the patch),
or do that one layer higher - don't modify `BenchmarkRunner::runConfiguration()`,
produce multiple reports, don't actually print each one, but aggregate them somehow
and only print the final one.
Initially i've gone ahead with the latter approach, but it didn't look like a natural fit;
the former (as in the diff) does seem like a better fit to me.
There's also a question of the test coverage. It sure currently does work here:
```
$ ./bin/llvm-exegesis --opcode-name=CMOV64rr --mode=inverse_throughput --repetition-mode=duplicate
Check generated assembly with: /usr/bin/objdump -d /tmp/snippet-8fb949.o
---
mode: inverse_throughput
key:
instructions:
- 'CMOV64rr RAX RAX R11 i_0x0'
- 'CMOV64rr RBP RBP R15 i_0x0'
- 'CMOV64rr RBX RBX RBX i_0x0'
- 'CMOV64rr RCX RCX RBX i_0x0'
- 'CMOV64rr RDI RDI R10 i_0x0'
- 'CMOV64rr RDX RDX RAX i_0x0'
- 'CMOV64rr RSI RSI RAX i_0x0'
- 'CMOV64rr R8 R8 R8 i_0x0'
- 'CMOV64rr R9 R9 RDX i_0x0'
- 'CMOV64rr R10 R10 RBX i_0x0'
- 'CMOV64rr R11 R11 R14 i_0x0'
- 'CMOV64rr R12 R12 R9 i_0x0'
- 'CMOV64rr R13 R13 R12 i_0x0'
- 'CMOV64rr R14 R14 R15 i_0x0'
- 'CMOV64rr R15 R15 R13 i_0x0'
config: ''
register_initial_values:
- 'RAX=0x0'
- 'R11=0x0'
- 'EFLAGS=0x0'
- 'RBP=0x0'
- 'R15=0x0'
- 'RBX=0x0'
- 'RCX=0x0'
- 'RDI=0x0'
- 'R10=0x0'
- 'RDX=0x0'
- 'RSI=0x0'
- 'R8=0x0'
- 'R9=0x0'
- 'R14=0x0'
- 'R12=0x0'
- 'R13=0x0'
cpu_name: bdver2
llvm_triple: x86_64-unknown-linux-gnu
num_repetitions: 10000
measurements:
- { key: inverse_throughput, value: 0.819, per_snippet_value: 12.285 }
error: ''
info: instruction has tied variables, using static renaming.
assembled_snippet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
...
$ ./bin/llvm-exegesis --opcode-name=CMOV64rr --mode=inverse_throughput --repetition-mode=loop
Check generated assembly with: /usr/bin/objdump -d /tmp/snippet-051eb3.o
---
mode: inverse_throughput
key:
instructions:
- 'CMOV64rr RAX RAX R11 i_0x0'
- 'CMOV64rr RBP RBP RSI i_0x0'
- 'CMOV64rr RBX RBX R9 i_0x0'
- 'CMOV64rr RCX RCX RSI i_0x0'
- 'CMOV64rr RDI RDI RBP i_0x0'
- 'CMOV64rr RDX RDX R9 i_0x0'
- 'CMOV64rr RSI RSI RDI i_0x0'
- 'CMOV64rr R9 R9 R12 i_0x0'
- 'CMOV64rr R10 R10 R11 i_0x0'
- 'CMOV64rr R11 R11 R9 i_0x0'
- 'CMOV64rr R12 R12 RBP i_0x0'
- 'CMOV64rr R13 R13 RSI i_0x0'
- 'CMOV64rr R14 R14 R14 i_0x0'
- 'CMOV64rr R15 R15 R10 i_0x0'
config: ''
register_initial_values:
- 'RAX=0x0'
- 'R11=0x0'
- 'EFLAGS=0x0'
- 'RBP=0x0'
- 'RSI=0x0'
- 'RBX=0x0'
- 'R9=0x0'
- 'RCX=0x0'
- 'RDI=0x0'
- 'RDX=0x0'
- 'R12=0x0'
- 'R10=0x0'
- 'R13=0x0'
- 'R14=0x0'
- 'R15=0x0'
cpu_name: bdver2
llvm_triple: x86_64-unknown-linux-gnu
num_repetitions: 10000
measurements:
- { key: inverse_throughput, value: 0.6083, per_snippet_value: 8.5162 }
error: ''
info: instruction has tied variables, using static renaming.
assembled_snippet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
...
$ ./bin/llvm-exegesis --opcode-name=CMOV64rr --mode=inverse_throughput --repetition-mode=min
Check generated assembly with: /usr/bin/objdump -d /tmp/snippet-c7a47d.o
Check generated assembly with: /usr/bin/objdump -d /tmp/snippet-2581f1.o
---
mode: inverse_throughput
key:
instructions:
- 'CMOV64rr RAX RAX R11 i_0x0'
- 'CMOV64rr RBP RBP R10 i_0x0'
- 'CMOV64rr RBX RBX R10 i_0x0'
- 'CMOV64rr RCX RCX RDX i_0x0'
- 'CMOV64rr RDI RDI RAX i_0x0'
- 'CMOV64rr RDX RDX R9 i_0x0'
- 'CMOV64rr RSI RSI RAX i_0x0'
- 'CMOV64rr R9 R9 RBX i_0x0'
- 'CMOV64rr R10 R10 R12 i_0x0'
- 'CMOV64rr R11 R11 RDI i_0x0'
- 'CMOV64rr R12 R12 RDI i_0x0'
- 'CMOV64rr R13 R13 RDI i_0x0'
- 'CMOV64rr R14 R14 R9 i_0x0'
- 'CMOV64rr R15 R15 RBP i_0x0'
config: ''
register_initial_values:
- 'RAX=0x0'
- 'R11=0x0'
- 'EFLAGS=0x0'
- 'RBP=0x0'
- 'R10=0x0'
- 'RBX=0x0'
- 'RCX=0x0'
- 'RDX=0x0'
- 'RDI=0x0'
- 'R9=0x0'
- 'RSI=0x0'
- 'R12=0x0'
- 'R13=0x0'
- 'R14=0x0'
- 'R15=0x0'
cpu_name: bdver2
llvm_triple: x86_64-unknown-linux-gnu
num_repetitions: 10000
measurements:
- { key: inverse_throughput, value: 0.6073, per_snippet_value: 8.5022 }
error: ''
info: instruction has tied variables, using static renaming.
assembled_snippet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
...
```
but i open to suggestions as to how test that.
I also have gone with the suggestion to default to this new mode.
This was irking me for some time, so i'm happy to finally see progress here.
Looking forward to feedback.
Reviewers: courbet, gchatelet
Reviewed By: courbet, gchatelet
Subscribers: mstojanovic, RKSimon, llvm-commits, courbet, gchatelet
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76921
The requirement for deopt parameter to be in gc parameter if it can
be modified by GC is very strong and difficult to follow.
The key example of why this can't work:
%p1 = bitcast i8* %p to i8*
statepoint [gc = (%p1)], [deopt = (%p1)]
The optimizer is allowed to replace either use (or both) of %p1 with %p.
If it updates only one of the two (entirely legal), the two sets do not overlap.
So this change removes the strong wording.
Reviewers: reames, dantrushin
Reviewed By: reames
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D77122
We already mention that `noalias` is modeled after the C99 `restrict`
qualifier but we did omit one important requirement in the description.
For the restrict guarantees the object affected has to be modified
during the execution of the function, in any way (see 6.7.3.1.4 in [0]).
There are two reasons we want this restriction as well:
1) To match the `restrict` semantics when we lower it to `noalias`.
2) To allow the reasoning that the object pointed to by a `noalias`
pointer is not modified through means not derived from this
pointer. Hence, following the uses of that pointer is sufficient
to determine potential modifications.
The discussion on this came up as part of D73428. In that patch the
Attributor is taught to derive `noalias` for call site arguments based
on alias queries against objects that are accessed in the callee. This
is possible even if the pointer passed at the call site was "not-`noalias`".
To simplify the logic there *and* to allow the use of `noalias` as
described in 2) above, it is beneficial to follow the C `restrict`
semantics in cases where there might be "read-read-aliases". Note that
AliasAnalysis* queries for read only objects already result in
`NoAlias` even if the pointers might "alias".
* From this point of view our Alias Analysis is basically a Dependence
Analysis.
[0] http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1124.pdf
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D74935
Summary:
This patch clarifies the semantics of branching on undef value.
Defining `br undef` as undefined behavior explains optimizations that use branch conditions, such as CVP (D76931) and GVN (propagateEquality).
For `switch cond`, it is defined to raise UB if cond is an expression containing undef && cond is not frozen &&
it may yield different values.
This allows that at the destination block the branch condition can be assumed to be frozen already (otherwise UB was already triggered).
This condition is slightly stricter than MemorySanitizer, which allows undef-y condition if it always leads to the same destination,
but it does not break MemorySanitizer because we are giving stricter constraint.
Reviewers: efriedma, fhahn, nikic, spatel, jdoerfert, nlopes
Reviewed By: nlopes
Subscribers: regehr, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76973
I added a list of options to configure should someone have issues with
long build time or running out of memory. This was added under common
problems in the getting started section of the documentation.
Reviewed By: Meinersbur, dim, e-leclercq
Differential Revision: https://reviews.llvm.org/D75425
This allows defining substitutions in terms of other substitutions. For
example, a %build substitution could be defined in terms of a %cxx
substitution as '%cxx %s -o %t.exe' and the script would be properly
expanded.
Differential Revision: https://reviews.llvm.org/D76178
1.Add instructions to update author when committing other's patch
We have updated DeveloperPolicy to show how to change author in
https://reviews.llvm.org/D72468
We should also update Phabricator page to include such infomation,
in case people follow the steps here and forget to update author info.
2. Replace `git llvm push` with `git push`
Reviewed By: probinson
Differential Revision: https://reviews.llvm.org/D76718
There has been some ongoing confusion regarding when to use `llvm_unreachable`
which this patch attempts to address. Specifically, the confusion has been
around whether `llvm_unreachable` is intended to mark only unreachable code
paths that the compiler cannot determine itself or to mark a code path which is
unconditionally a bug to reach. Based on email and IRC discussions, it sounds
like "unconditional bug to reach" is the consensus.
to remap object file paths (but no source paths) before
processing. This is meant to be used for Clang objects where the
module cache location was remapped using ``-fdebug-prefix-map``; to
help dsymutil find the Clang module cache.
<rdar://problem/55685132>
Differential Revision: https://reviews.llvm.org/D76391
Summary:
The next release of LLVM will support the full ACLE spec for MVE intrinsics,
so it's worth saying so in the release notes.
Reviewers: kristof.beyls
Reviewed By: kristof.beyls
Subscribers: cfe-commits, hans, dmgreen, llvm-commits
Tags: #llvm, #clang
Differential Revision: https://reviews.llvm.org/D76513
Summary:
Add new generic MIR opcodes G_SADDSAT etc. Add support in IRTranslator
for translating the saturating add/subtract intrinsics to the new
opcodes.
Reviewers: aemerson, dsanders, paquette, arsenm
Subscribers: jvesely, wdng, nhaehnle, rovka, hiraditya, volkan, kerbowa, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D76600
This handles not paths embedded in debug info, but also in sources.
Since the use of this flag is controlled by an option, rather than
replacing the new option, we add a new option.
Differential Revision: https://reviews.llvm.org/D76018
Remove the gap left between the stack pointer (s32) and frame pointer
(s34) now that the scratch wave offset is no longer a part of the
calling convention ABI.
Update llvm/docs/AMDGPUUsage.rst to reflect the change.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75657
Add the scratch wave offset to the scratch buffer descriptor (SRSrc) in
the entry function prologue. This allows us to removes the scratch wave
offset register from the calling convention ABI.
As part of this change, allow the use of an inline constant zero for the
SOffset of MUBUF instructions accessing the stack in entry functions
when a frame pointer is not requested/required. Entry functions with
calls still need to set up the calling convention ABI stack pointer
register, and reference it in order to address arguments of called
functions. The ABI stack pointer register remains unswizzled, but is now
wave-relative instead of queue-relative.
Non-entry functions also use an inline constant zero SOffset for
wave-relative scratch access, but continue to use the stack and frame
pointers as before. When the stack or frame pointer is converted to a
swizzled offset it is now scaled directly, as the scratch wave offset no
longer needs to be subtracted first.
Update llvm/docs/AMDGPUUsage.rst to reflect these changes to the calling
convention.
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75138
This is copied from the suggested text by @regehr in:
https://bugs.llvm.org/show_bug.cgi?id=20895
The way forward was not clear for several years, but now that we
have 'freeze' and Alive2, the behavior should be documented.
Also see comments in D76332.
Makes tests fail on Windows, see https://reviews.llvm.org/D70720#1924542
This reverts commit 3a5ddedadb671e485ce5c638142817879ac14a8c, and
follow-ups:
f4cb9c919e28276222873453cf85de9e5a3c7be5
042eb0482aa758057c4f77616a4696cdb21b4fcc
c0cf5f5da9a7bf1bdf43ed53287b0f634fc53045
18649f48139932377c2a2909f1fb600bf5cf6e57
f62b898c1f5dd77e68b53570dc2679877bcbe4c2
This adds the --debug-vars option to llvm-objdump, which prints
locations (registers/memory) of source-level variables alongside the
disassembly based on DWARF info. A vertical line is printed for each
live-range, with a label at the top giving the variable name and
location, and the position and length of the line indicating the program
counter range in which it is valid.
Currently, this only works for object files, not executables or shared
libraries.
Differential revision: https://reviews.llvm.org/D70720
AVR has been enabled by default since
c480c584a0b7de675dddb2616122fc218cd72c0e, the tests have been stable for
a couple days now, revert extremely unlikely.
LLVM currently supports CSK_MD5 and CSK_SHA1 source file checksums in
debug info. This change adds support for CSK_SHA256 checksums.
The SHA256 checksums are supported by the CodeView debug format.
Reviewed By: aprantl
Differential Revision: https://reviews.llvm.org/D75785
Summary: This patch adds the basic utilities to deal with dropable uses. dropable uses are uses that we rather drop than prevent transformations, for now they are limited to uses in llvm.assume.
Reviewers: jdoerfert, sstefan1
Reviewed By: jdoerfert
Subscribers: uenoku, lebedev.ri, mgorny, hiraditya, dexonsmith, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D73404
Summary: As discussed in D75505, it's not particularly useful to change the type of a load to/from floating-point/integer because it's followed by a bitcast, and it might lead to surprising code generation. Check that this doesn't generally happen.
Reviewers: lebedev.ri
Subscribers: jkorous, dexonsmith, ributzka, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D75644
This is an update to the documentation of our community code-review process.
Based on the RFC: High-Level Code-Review Documentation Update
(http://lists.llvm.org/pipermail/llvm-dev/2019-November/136808.html).
In this patch, I've pulled out the documentation into a separate file, and
broken it into a number of subsections. This is, of course, just one further
step in better documenting our community processes. I expect we'll continue to
improve this over time. Thank you to everyone who provided feedback!
Differential Revision: https://reviews.llvm.org/D71916
