Summary:
Remove explicit stages and introduce a list of stages.
A pipeline should be composed of an arbitrary list of stages, and not any
predefined list of stages in the Backend. The Backend should not know of any
particular stage, rather it should only be concerned that it has a list of
stages, and that those stages will fulfill the contract of what it means to be
a Stage (namely pre/post/execute a given instruction).
For now, we leave the original set of stages defined in the Backend ctor;
however, I imagine these will be moved out at a later time.
This patch makes an adjustment to the semantics of Stage::isReady.
Specifically, what the Backend really needs to know is if a Stage has
unfinished work. With that said, it is more appropriately renamed
Stage::hasWorkToComplete(). This change will clean up the check in
Backend::run(), allowing us to query each stage to see if there is unfinished
work, regardless of what subclass a stage might be. I feel that this change
simplifies the semantics too, but that's a subjective statement.
Given how RetireStage and ExecuteStage handle data in their preExecute(), I've
had to change the order of Retire and Execute in our stage list. Retire must
complete any of its preExecute actions before ExecuteStage's preExecute can
take control. This is mainly because both stages utilize the RCU. In the
meantime, I want to see if I can adjust that or remove that coupling.
Reviewers: andreadb, RKSimon, courbet
Reviewed By: andreadb
Subscribers: tschuett, gbedwell, llvm-commits
Differential Revision: https://reviews.llvm.org/D46907
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@335361 91177308-0d34-0410-b5e6-96231b3b80d8
After the recent refactoring that introduced parallel handling of
different object, the binary holder became unique per object file. This
defeats its optimization of caching archives, leading to an archive
being opened for every binary it contains. This is obviously unfortunate
and will need to be refactored soon.
Luckily in practice, the impact of this is limited as most files are
mmap'ed instead of memcopy'd. There's a caveat however: when the memory
buffer requires a null terminator and it's a multiple of the page size,
we allocate instead of mmap'ing. If this happens for a static archive,
we end up with N copies of it in memory, where N is the number of
objects in the archive, leading to exuberant memory usage. This provided
a stopgap solution to ensure that all the files it loads are mmap in
memory by removing the requirement for a terminating null byte.
Differential revision: https://reviews.llvm.org/D48397
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@335293 91177308-0d34-0410-b5e6-96231b3b80d8
Also, rename fields `TotalMappings` and `NumUsedMappings` in struct
RegisterMappingTracker into `NumPhysRegs` and `NumUsedPhysRegs`.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@335219 91177308-0d34-0410-b5e6-96231b3b80d8
This patch teaches llvm-mca how to identify register writes that implicitly zero
the upper portion of a super-register.
On X86-64, a general purpose register is implemented in hardware as a 64-bit
register. Quoting the Intel 64 Software Developer's Manual: "an update to the
lower 32 bits of a 64 bit integer register is architecturally defined to zero
extend the upper 32 bits". Also, a write to an XMM register performed by an AVX
instruction implicitly zeroes the upper 128 bits of the aliasing YMM register.
This patch adds a new method named clearsSuperRegisters to the MCInstrAnalysis
interface to help identify instructions that implicitly clear the upper portion
of a super-register. The rest of the patch teaches llvm-mca how to use that new
method to obtain the information, and update the register dependencies
accordingly.
I compared the kernels from tests clear-super-register-1.s and
clear-super-register-2.s against the output from perf on btver2. Previously
there was a large discrepancy between the estimated IPC and the measured IPC.
Now the differences are mostly in the noise.
Differential Revision: https://reviews.llvm.org/D48225
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@335113 91177308-0d34-0410-b5e6-96231b3b80d8
Summary: Introducing a Prototype object to capture Variables that must be set but keeps degrees of freedom as Invalid. This allows exploring non constraint variables later on.
Reviewers: courbet
Subscribers: tschuett, llvm-commits
Differential Revision: https://reviews.llvm.org/D48316
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@335105 91177308-0d34-0410-b5e6-96231b3b80d8
Histogram entries are now ordered by key. This should improves their
readability when statistics are printed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334961 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
While that is indeed a quite interesting summary stat,
there are cases where it does not really add anything
other than consuming extra lines.
Declutters the output of D48190.
Reviewers: RKSimon, andreadb, courbet, craig.topper
Reviewed By: andreadb
Subscribers: javed.absar, gbedwell, llvm-commits
Differential Revision: https://reviews.llvm.org/D48209
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334833 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
On hover, the whole asm snippet is displayed, including operands.
This requires the actual assembly output instead of just the MCInsts:
This is because some pseudo-instructions get lowered to actual target
instructions during codegen (e.g. ABS_Fp32 -> SSE or X87).
Reviewers: gchatelet
Subscribers: mgorny, tschuett, llvm-commits
Differential Revision: https://reviews.llvm.org/D48164
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334805 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Get rid of OpcodeName.
To remove the opcode name from an old file:
```
cat old_file | sed '/opcode_name.*/d'
```
Reviewers: gchatelet
Subscribers: tschuett, llvm-commits
Differential Revision: https://reviews.llvm.org/D48121
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334691 91177308-0d34-0410-b5e6-96231b3b80d8
Summary: This patch transforms the Scheduler class into the ExecuteStage. Most of the logic remains.
Reviewers: andreadb, RKSimon, courbet
Reviewed By: andreadb
Subscribers: mgorny, javed.absar, tschuett, gbedwell, llvm-commits
Differential Revision: https://reviews.llvm.org/D47246
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334679 91177308-0d34-0410-b5e6-96231b3b80d8
Fixes PR37790.
In some (very rare) cases, the LSUnit (Load/Store unit) was wrongly marking a
load (or store) as "ready to execute" effectively bypassing older memory barrier
instructions.
To reproduce this bug, the memory barrier must be the first instruction in the
input assembly sequence, and it doesn't have to perform any register writes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334633 91177308-0d34-0410-b5e6-96231b3b80d8
Not sure why, but it breaks buildbot clang-cmake-armv8-full.
It causes a failure in TEST 'Xray-armhf-linux :: TestCases/Posix/profiling-single-threaded.cc'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334617 91177308-0d34-0410-b5e6-96231b3b80d8
Summary: Previous design was relying on the 'mutate' keyword and was quite confusing. This version separate mutable from immutable data and makes it clearer what changes and what doesn't.
Reviewers: courbet
Subscribers: tschuett, llvm-commits
Differential Revision: https://reviews.llvm.org/D48020
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334596 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
This method was not correct for entries in DWO files as it assumed it
could just add up the CU and DIE offsets to get the absolute DIE offset.
This is not correct for the DWO files, as here the CU offset will
reference the skeleton unit, whereas the DIE offset will be the offset
in the full unit in the DWO file.
Unfortunately, this means that we are not able to determine the absolute
DIE offset using the information in the .debug_names section alone,
which means we have to offload some of this work to the users of this
class.
To demonstrate how this can be done, I've added/fixed the ability to
lookup entries using accelerator tables in DWO files in llvm-dwarfdump.
To make this happen, I've needed to make two extra changes in other
classes:
- made the DWARFContext method to lookup a CU based on the section
offset public. I've needed this functionality to lookup a CU, and this
seems like a useful thing in general.
- made DWARFUnit::getDWOId call extractDIEsIfNeeded. Before this, the
DWOId was filled in only if the root DIE happened to be parsed
before we called the accessor. Since the lazy parsing is supposed to
happen under the hood, calling extractDIEsIfNeeded seems appropriate.
Reviewers: JDevlieghere, aprantl, dblaikie
Subscribers: mgrang, llvm-commits
Differential Revision: https://reviews.llvm.org/D48009
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334578 91177308-0d34-0410-b5e6-96231b3b80d8
This simplifies some code which had StringRefs to begin with, and
makes other code more complicated which had const char* to begin
with.
In the end, I think this makes for a more idiomatic and platform
agnostic API. Not all platforms launch process with null terminated
c-string arrays for the environment pointer and argv, but the api
was designed that way because it allowed easy pass-through for
posix-based platforms. There's a little additional overhead now
since on posix based platforms we'll be takign StringRefs which
were constructed from null terminated strings and then copying
them to null terminate them again, but from a readability and
usability standpoint of the API user, I think this API signature
is strictly better.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334518 91177308-0d34-0410-b5e6-96231b3b80d8
Don't provide the assembler source as the "root file" unless the user
asked to have debug info for the assembler source (with -g).
If the source doesn't provide an explicit ".file 0" then (a) use the
compilation directory as directory #0, and (b) use the file #1 info
for file #0 also.
Differential Revision: https://reviews.llvm.org/D48055
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334512 91177308-0d34-0410-b5e6-96231b3b80d8
Some out-of-tree targets depend on the enum name SPF_Binary. Keep the name
can avoid unnecessary churn to those targets.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334476 91177308-0d34-0410-b5e6-96231b3b80d8
Name table occupies a big chunk of size in current binary format sample profile.
In order to reduce its size, the patch changes the sample writer/reader to
save/restore MD5Hash of names in the name table. Sample annotation phase will
also use MD5Hash of name to query samples accordingly.
Experiment shows compact binary format can reduce the size of sample profile by
2/3 compared with binary format generally.
Differential Revision: https://reviews.llvm.org/D47955
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334447 91177308-0d34-0410-b5e6-96231b3b80d8
NFC here, this just raises some platform specific ifdef hackery
out of a class and creates proper platform-independent typedefs
for the relevant things. This allows these typedefs to be
reused in other places without having to reinvent this preprocessor
logic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334294 91177308-0d34-0410-b5e6-96231b3b80d8
I don't know how to build this code, but based on the failing
buildbot error message it looks like this change should get
the buildbot up and running again.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334231 91177308-0d34-0410-b5e6-96231b3b80d8
This breaks the OpenFlags enumeration into two separate
enumerations: OpenFlags and CreationDisposition. The first
controls the behavior of the API depending on whether or not
the target file already exists, and is not a flags-based
enum. The second controls more flags-like values.
This yields a more easy to understand API, while also allowing
flags to be passed to the openForRead api, where most of the
values didn't make sense before. This also makes the apis more
testable as it becomes easy to enumerate all the configurations
which make sense, so I've added many new tests to exercise all
the different values.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334221 91177308-0d34-0410-b5e6-96231b3b80d8
The class Object contains std::shared_ptr<MemoryBuffer> OwnedData
which is not used anywhere. Besides avoiding two stage initialization
the motivation to remove it comes from the plan to add (currently missing) support
for static libraries.
NFC.
Test plan: make check-all
Differential revision: https://reviews.llvm.org/D47855
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@334217 91177308-0d34-0410-b5e6-96231b3b80d8
Andrea Di Biagio