Since this can be set with s_setreg*, it should not be a subtarget
property. Set a default based on the calling convention, and Introduce
a new amdgpu-dx10-clamp attribute to override this if desired.
Also introduce a new amdgpu-ieee attribute to match.
The values need to match to allow inlining. I think it is OK for the
caller's dx10-clamp attribute to override the callee, but there
doesn't appear to be the infrastructure to do this currently without
definining the attribute in the generic Attributes.td.
Eventually the calling convention lowering will need to insert a mode
switch somewhere for these.
llvm-svn: 357302
Document the intended use of the `.amdgcn.next_free_{s,v}gpr` in the
context of multiple kernels and functions.
Differential Revision: https://reviews.llvm.org/D59949
llvm-svn: 357289
Ensure Code Object V2 documentation is complete, but always contains a
warning and a link to the equivalent Code Object V3 documentation.
Explicitly indicate that any note records present in a code object that
are not documented must be considered deprecated and ignored.
Differential Revision: https://reviews.llvm.org/D59782
llvm-svn: 357176
Summary:
This is an alternative to D59539.
Let's suppose we have measured 4 different opcodes, and got: `0.5`, `1.0`, `1.5`, `2.0`.
Let's suppose we are using `-analysis-clustering-epsilon=0.5`.
By default now we will start processing the `0.5` point, find that `1.0` is it's neighbor, add them to a new cluster.
Then we will notice that `1.5` is a neighbor of `1.0` and add it to that same cluster.
Then we will notice that `2.0` is a neighbor of `1.5` and add it to that same cluster.
So all these points ended up in the same cluster.
This may or may not be a correct implementation of dbscan clustering algorithm.
But this is rather horribly broken for the reasons of comparing the clusters with the LLVM sched data.
Let's suppose all those opcodes are currently in the same sched cluster.
If i specify `-analysis-inconsistency-epsilon=0.5`, then no matter
the LLVM values this cluster will **never** match the LLVM values,
and thus this cluster will **always** be displayed as inconsistent.
The solution is obviously to split off some of these opcodes into different sched cluster.
But how do i do that? Out of 4 opcodes displayed in the inconsistency report,
which ones are the "bad ones"? Which ones are the most different from the checked-in data?
I'd need to go in to the `.yaml` and look it up manually.
The trivial solution is to, when creating clusters, don't use the full dbscan algorithm,
but instead "pick some unclustered point, pick all unclustered points that are it's neighbor,
put them all into a new cluster, repeat". And just so as it happens, we can arrive
at that algorithm by not performing the "add neighbors of a neighbor to the cluster" step.
But that won't work well once we teach analyze mode to operate in on-1D mode
(i.e. on more than a single measurement type at a time), because the clustering would
depend on the order of the measurements.
Instead, let's just create a single cluster per opcode, and put all the points of that opcode into said cluster.
And simultaneously check that every point in that cluster is a neighbor of every other point in the cluster,
and if they are not, the cluster (==opcode) is unstable.
This is //yet another// step to bring me closer to being able to continue cleanup of bdver2 sched model..
Fixes [[ https://bugs.llvm.org/show_bug.cgi?id=40880 | PR40880 ]].
Reviewers: courbet, gchatelet
Reviewed By: courbet
Subscribers: tschuett, jdoerfert, RKSimon, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59820
llvm-svn: 357152
A section containing metadata on remark diagnostics will be emitted if
the flag (-mllvm) -remarks-section is present.
For now, the metadata is:
* a magic number for remarks: "REMARKS\0"
* the version number: a little-endian uint64_t
* the absolute file path to the serialized remark diagnostics: a
null-terminated string.
Differential Revision: https://reviews.llvm.org/D59571
llvm-svn: 357043
Just as as llvm IR supports explicitly specifying numeric value ids
for instructions, and emits them by default in textual output, now do
the same for blocks.
This is a slightly incompatible change in the textual IR format.
Previously, llvm would parse numeric labels as string names. E.g.
define void @f() {
br label %"55"
55:
ret void
}
defined a label *named* "55", even without needing to be quoted, while
the reference required quoting. Now, if you intend a block label which
looks like a value number to be a name, you must quote it in the
definition too (e.g. `"55":`).
Previously, llvm would print nameless blocks only as a comment, and
would omit it if there was no predecessor. This could cause confusion
for readers of the IR, just as unnamed instructions did prior to the
addition of "%5 = " syntax, back in 2008 (PR2480).
Now, it will always print a label for an unnamed block, with the
exception of the entry block. (IMO it may be better to print it for
the entry-block as well. However, that requires updating many more
tests.)
Thus, the following is supported, and is the canonical printing:
define i32 @f(i32, i32) {
%3 = add i32 %0, %1
br label %4
4:
ret i32 %3
}
New test cases covering this behavior are added, and other tests
updated as required.
Differential Revision: https://reviews.llvm.org/D58548
llvm-svn: 356789
Introduce a DW_OP_LLVM_convert Dwarf expression pseudo op that allows
for a convenient way to perform type conversions on the Dwarf expression
stack. As an additional bonus it paves the way for using other Dwarf
v5 ops that need to reference a base_type.
The new DW_OP_LLVM_convert is used from lib/Transforms/Utils/Local.cpp
to perform sext/zext on debug values but mainly the patch is about
preparing terrain for adding other Dwarf v5 ops that need to reference a
base_type.
For Dwarf v5 the op maps to DW_OP_convert and for earlier versions a
complex shift & mask pattern is generated to emulate sext/zext.
This is a recommit of r356442 with trivial fixes for the failing tests.
Differential Revision: https://reviews.llvm.org/D56587
llvm-svn: 356451
Introduce a DW_OP_LLVM_convert Dwarf expression pseudo op that allows
for a convenient way to perform type conversions on the Dwarf expression
stack. As an additional bonus it paves the way for using other Dwarf
v5 ops that need to reference a base_type.
The new DW_OP_LLVM_convert is used from lib/Transforms/Utils/Local.cpp
to perform sext/zext on debug values but mainly the patch is about
preparing terrain for adding other Dwarf v5 ops that need to reference a
base_type.
For Dwarf v5 the op maps to DW_OP_convert and for earlier versions a
complex shift & mask pattern is generated to emulate sext/zext.
Differential Revision: https://reviews.llvm.org/D56587
llvm-svn: 356442
Add an experimental buffer fat pointer address space that is currently
unhandled in the backend. This commit reserves address space 7 as a
non-integral pointer repsenting the 160-bit fat pointer (128-bit buffer
descriptor + 32-bit offset) that is heavily used in graphics workloads
using the AMDGPU backend.
Differential Revision: https://reviews.llvm.org/D58957
llvm-svn: 356373
Summary:
CoverageExporterJson::renderFiles accounts for most of the execution time given a large profdata file with multiple binaries.
Proposed solution is to generate JSON for each file in parallel and sort at the end to preserve deterministic output. Also added flags to skip generating parts of the output to trim the output size.
Patch by Sajjad Mirza (@sajjadm).
Reviewers: Dor1s, vsk
Reviewed By: Dor1s, vsk
Subscribers: liaoyuke, mgrang, jdoerfert, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D59277
llvm-svn: 356178
This patch adds a section, ``Object lifetime in optimized code'', that
documents how such intrinsics are supposed to be handled. It sets out some of
the principles of how they specify variable locations, and how long those
locations are valid for.
This patch also documents one of the objectives behind the variable-location
design, that we should never allow the debugger to observe a state of the
program that would not have appeared without optimization.
Differential Revision: https://reviews.llvm.org/D58726
llvm-svn: 356041
This indicates an intrinsic parameter is required to be a constant,
and should not be replaced with a non-constant value.
Add the attribute to all AMDGPU and generic intrinsics that comments
indicate it should apply to. I scanned other target intrinsics, but I
don't see any obvious comments indicating which arguments are intended
to be only immediates.
This breaks one questionable testcase for the autoupgrade. I'm unclear
on whether the autoupgrade is supposed to really handle declarations
which were never valid. The verifier fails because the attributes now
refer to a parameter past the end of the argument list.
llvm-svn: 355981
Add a note about legacy FunctionPassManager to the LLVM tutorial.
It seems to confuse some people, worth adding a warning to the tutorial
to elaborate and suggest using `llvm::legacy::FunctionPassManager` for
now. Not a perfect solution but hopefully will avoid confusion
in the meantime.
Differential Revision: https://reviews.llvm.org/D59258
llvm-svn: 355930
Use this feature to fix a bug on ARM where 4 byte alignment is
incorrectly assumed.
Differential Revision: https://reviews.llvm.org/D57335
llvm-svn: 355685
Summary:
Right now, when we encounter a string equality check,
e.g. `if (memcmp(a, b, s) == 0)`, we try to expand to a comparison if `s` is a
small compile-time constant, and fall back on calling `memcmp()` else.
This is sub-optimal because memcmp has to compute much more than
equality.
This patch replaces `memcmp(a, b, s) == 0` by `bcmp(a, b, s) == 0` on platforms
that support `bcmp`.
`bcmp` can be made much more efficient than `memcmp` because equality
compare is trivially parallel while lexicographic ordering has a chain
dependency.
Subscribers: fedor.sergeev, jyknight, ckennelly, gchatelet, llvm-commits
Differential Revision: https://reviews.llvm.org/D56593
llvm-svn: 355672
Use this feature to fix a bug on ARM where 4 byte alignment is
incorrectly assumed.
Differential Revision: https://reviews.llvm.org/D57335
llvm-svn: 355585
Use this feature to fix a bug on ARM where 4 byte alignment is
incorrectly assumed.
Differential Revision: https://reviews.llvm.org/D57335
llvm-svn: 355522
Summary:
The current install-clang-headers target installs clang's resource
directory headers. This is different from the install-llvm-headers
target, which installs LLVM's API headers. We want to introduce the
corresponding target to clang, and the natural name for that new target
would be install-clang-headers. Rename the existing target to
install-clang-resource-headers to free up the install-clang-headers name
for the new target, following the discussion on cfe-dev [1].
I didn't find any bots on zorg referencing install-clang-headers. I'll
send out another PSA to cfe-dev to accompany this rename.
[1] http://lists.llvm.org/pipermail/cfe-dev/2019-February/061365.html
Reviewers: beanz, phosek, tstellar, rnk, dim, serge-sans-paille
Subscribers: mgorny, javed.absar, jdoerfert, #sanitizers, openmp-commits, lldb-commits, cfe-commits, llvm-commits
Tags: #clang, #sanitizers, #lldb, #openmp, #llvm
Differential Revision: https://reviews.llvm.org/D58791
llvm-svn: 355340
This patch adds a new flag named -bottleneck-analysis to print out information
about throughput bottlenecks.
MCA knows how to identify and classify dynamic dispatch stalls. However, it
doesn't know how to analyze and highlight kernel bottlenecks. The goal of this
patch is to teach MCA how to correlate increases in backend pressure to backend
stalls (and therefore, the loss of throughput).
From a Scheduler point of view, backend pressure is a function of the scheduler
buffer usage (i.e. how the number of uOps in the scheduler buffers changes over
time). Backend pressure increases (or decreases) when there is a mismatch
between the number of opcodes dispatched, and the number of opcodes issued in
the same cycle. Since buffer resources are limited, continuous increases in
backend pressure would eventually leads to dispatch stalls. So, there is a
strong correlation between dispatch stalls, and how backpressure changed over
time.
This patch teaches how to identify situations where backend pressure increases
due to:
- unavailable pipeline resources.
- data dependencies.
Data dependencies may delay execution of instructions and therefore increase the
time that uOps have to spend in the scheduler buffers. That often translates to
an increase in backend pressure which may eventually lead to a bottleneck.
Contention on pipeline resources may also delay execution of instructions, and
lead to a temporary increase in backend pressure.
Internally, the Scheduler classifies instructions based on whether register /
memory operands are available or not.
An instruction is marked as "ready to execute" only if data dependencies are
fully resolved.
Every cycle, the Scheduler attempts to execute all instructions that are ready
to execute. If an instruction cannot execute because of unavailable pipeline
resources, then the Scheduler internally updates a BusyResourceUnits mask with
the ID of each unavailable resource.
ExecuteStage is responsible for tracking changes in backend pressure. If backend
pressure increases during a cycle because of contention on pipeline resources,
then ExecuteStage sends a "backend pressure" event to the listeners.
That event would contain information about instructions delayed by resource
pressure, as well as the BusyResourceUnits mask.
Note that ExecuteStage also knows how to identify situations where backpressure
increased because of delays introduced by data dependencies.
The SummaryView observes "backend pressure" events and prints out a "bottleneck
report".
Example of bottleneck report:
```
Cycles with backend pressure increase [ 99.89% ]
Throughput Bottlenecks:
Resource Pressure [ 0.00% ]
Data Dependencies: [ 99.89% ]
- Register Dependencies [ 0.00% ]
- Memory Dependencies [ 99.89% ]
```
A bottleneck report is printed out only if increases in backend pressure
eventually caused backend stalls.
About the time complexity:
Time complexity is linear in the number of instructions in the
Scheduler::PendingSet.
The average slowdown tends to be in the range of ~5-6%.
For memory intensive kernels, the slowdown can be significant if flag
-noalias=false is specified. In the worst case scenario I have observed a
slowdown of ~30% when flag -noalias=false was specified.
We can definitely recover part of that slowdown if we optimize class LSUnit (by
doing extra bookkeeping to speedup queries). For now, this new analysis is
disabled by default, and it can be enabled via flag -bottleneck-analysis. Users
of MCA as a library can enable the generation of pressure events through the
constructor of ExecuteStage.
This patch partially addresses https://bugs.llvm.org/show_bug.cgi?id=37494
Differential Revision: https://reviews.llvm.org/D58728
llvm-svn: 355308
This is a small addition to arithmetic operations that improves
expressiveness of the language.
Differential Revision: https://reviews.llvm.org/D58775
llvm-svn: 355187
This patch allows all forms of values for options to be used at the end
of a group. With the fix, it is possible to follow the way GNU binutils
tools handle grouping options better. For example, the -j option can be
used with objdump in any of the following ways:
$ objdump -d -j .text a.o
$ objdump -d -j.text a.o
$ objdump -dj .text a.o
$ objdump -dj.text a.o
Differential Revision: https://reviews.llvm.org/D58711
llvm-svn: 355185
If an option, which requires a value, has a `cl::Grouping` formatting
modifier, it works well as far as it is used at the end of a group,
or as a separate argument. However, if the option appears accidentally
in the middle of a group, the program just crashes. This patch prints
an error message instead.
Differential Revision: https://reviews.llvm.org/D58499
llvm-svn: 355184
Part 2 of CSPGO changes (mostly related to ProfileSummary).
Note that I use a default parameter in setProfileSummary() and getSummary().
This is to break the dependency in clang. I will make the parameter explicit
after changing clang in a separated patch.
Differential Revision: https://reviews.llvm.org/D54175
llvm-svn: 355131
Fix https://bugs.llvm.org/show_bug.cgi?id=38583: Describe
how memcpy/memmove/memset behave when len=0. Also fix
some fallout from when the alignment parameter was
replaced by an attribute.
This closes PR38583.
Patch by RalfJung (Ralf)
Differential Revision: https://reviews.llvm.org/D57600
llvm-svn: 354911
We have all the necessary legalization, expansion and unrolling support required for the *.overflow intrinsics with vector types, so update the docs to make that clear.
Note: vectorization is not in place yet (the non-homogenous return types aren't well supported) so we still must explicitly use the vectors intrinsics and not reply on slp/loop.
Differential Revision: https://reviews.llvm.org/D58618
llvm-svn: 354821
The patch adds support for --hash-filenames to llvm-cov. This option adds md5
hash of the source path to the name of the generated .gcov file. The option is
crucial for cases where you have multiple files with the same name but can't
use --preserve-paths as resulting filenames exceed the limit.
from gcov(1):
```
-x
--hash-filenames
By default, gcov uses the full pathname of the source files to to
create an output filename. This can lead to long filenames that
can overflow filesystem limits. This option creates names of the
form source-file##md5.gcov, where the source-file component is
the final filename part and the md5 component is calculated from
the full mangled name that would have been used otherwise.
```
Patch by Igor Ignatev!
Differential Revision: https://reviews.llvm.org/D58370
llvm-svn: 354379
Matt Arsenault