The range should be assumed to be the hardware maximum
if a workitem intrinsic is used in a callable function
which does not know the restricted limit of the calling
kernel.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316346 91177308-0d34-0410-b5e6-96231b3b80d8
Should be no functional change for now. A future disassembler change will prevent disassembling with 0xf2/0xf3.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316339 91177308-0d34-0410-b5e6-96231b3b80d8
I don't think this changes anything functionally yet, but I plan to fix the disassembler to use this to disable matching certain instructions with 0xf3/0xf2/0x66 prefixes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316337 91177308-0d34-0410-b5e6-96231b3b80d8
combineShuffleOfScalars is very conservative about shuffled BUILD_VECTORs that can be combined together.
This patch adds one additional case - if both BUILD_VECTORs represent splats of the same scalar value but with different UNDEF elements, then we should create a single splat BUILD_VECTOR, sharing only the UNDEF elements defined by the shuffle mask.
Differential Revision: https://reviews.llvm.org/D38696
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316331 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Support formatting formatv_objects.
While here, fix documentation about member-formatters, and attempted
perfect-forwarding (I think).
Reviewers: zturner
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D38997
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316330 91177308-0d34-0410-b5e6-96231b3b80d8
Before, loop unrolling was only enabled for loops with a single
block. This restriction has been removed and replaced by:
- allow a maximum of two exiting blocks,
- a four basic block limit for cores with a branch predictor.
Differential Revision: https://reviews.llvm.org/D38952
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316313 91177308-0d34-0410-b5e6-96231b3b80d8
The overflow detection assertions were tautological due to truncation.
Adjust them to no longer be tautological.
Patch by Alex Langford!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316303 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
It's unclear if this is the only thing we can do but at least this is consistent with the check
of address space agreement in `isBitCastable`.
The code is used at least in both instcombine and jumpthreading though
I could only find a way to trigger the invalid cast in instcombine.
Reviewers: loladiro, sanjoy, majnemer
Reviewed By: sanjoy
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D34335
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316302 91177308-0d34-0410-b5e6-96231b3b80d8
As discussed in D39011:
https://reviews.llvm.org/D39011
...replacing constants with a variable is inverting the transform done
by other IR passes, so we definitely don't want to do this early.
In fact, it's questionable whether this transform belongs in SimplifyCFG
at all. I'll look at moving this to codegen as a follow-up step.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316298 91177308-0d34-0410-b5e6-96231b3b80d8
This fixes bugzilla 26810
https://bugs.llvm.org/show_bug.cgi?id=26810
This is intended to prevent sequences like:
movl %ebp, 8(%esp) # 4-byte Spill
movl %ecx, %ebp
movl %ebx, %ecx
movl %edi, %ebx
movl %edx, %edi
cltd
idivl %esi
movl %edi, %edx
movl %ebx, %edi
movl %ecx, %ebx
movl %ebp, %ecx
movl 16(%esp), %ebp # 4 - byte Reload
Such sequences are created in 2 scenarios:
Scenario #1:
vreg0 is evicted from physreg0 by vreg1
Evictee vreg0 is intended for region splitting with split candidate physreg0 (the reg vreg0 was evicted from)
Region splitting creates a local interval because of interference with the evictor vreg1 (normally region spliiting creates 2 interval, the "by reg" and "by stack" intervals. Local interval created when interference occurs.)
one of the split intervals ends up evicting vreg2 from physreg1
Evictee vreg2 is intended for region splitting with split candidate physreg1
one of the split intervals ends up evicting vreg3 from physreg2 etc.. until someone spills
Scenario #2
vreg0 is evicted from physreg0 by vreg1
vreg2 is evicted from physreg2 by vreg3 etc
Evictee vreg0 is intended for region splitting with split candidate physreg1
Region splitting creates a local interval because of interference with the evictor vreg1
one of the split intervals ends up evicting back original evictor vreg1 from physreg0 (the reg vreg0 was evicted from)
Another evictee vreg2 is intended for region splitting with split candidate physreg1
one of the split intervals ends up evicting vreg3 from physreg2 etc.. until someone spills
As compile time was a concern, I've added a flag to control weather we do cost calculations for local intervals we expect to be created (it's on by default for X86 target, off for the rest).
Differential Revision: https://reviews.llvm.org/D35816
Change-Id: Id9411ff7bbb845463d289ba2ae97737a1ee7cc39
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316295 91177308-0d34-0410-b5e6-96231b3b80d8
This is similar to how we generate the VEX tables.
More fixes are still needed for the instructions that use EVEX.b (broadcast and embedded rounding).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316294 91177308-0d34-0410-b5e6-96231b3b80d8
The missed canonicalization/optimization in the motivating test from PR34471 leads to very different codegen:
int switcher(int x) {
switch(x) {
case 17: return 17;
case 19: return 19;
case 42: return 42;
default: break;
}
return 0;
}
int comparator(int x) {
if (x == 17) return 17;
if (x == 19) return 19;
if (x == 42) return 42;
return 0;
}
For the first example, we use a bit-test optimization to avoid a series of compare-and-branch:
https://godbolt.org/g/BivDsw
Differential Revision: https://reviews.llvm.org/D39011
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316293 91177308-0d34-0410-b5e6-96231b3b80d8
This patch implements dynamic stack (re-)alignment for 16-bit Thumb. When
targeting processors, which support only the 16-bit Thumb instruction set
the compiler ignores the alignment attributes of automatic variables and may
silently generate incorrect code.
Differential revision: https://reviews.llvm.org/D38143
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316289 91177308-0d34-0410-b5e6-96231b3b80d8
The pass scans the function to find instruction chains that define
registers in the same domain (closures).
It then calculates the cost of converting the closure to another domain.
If found profitable, the instructions are converted to instructions in
the other domain and the register classes are changed accordingly.
This commit adds the pass infrastructure and a simple conversion from
the GPR domain to the Mask domain.
Differential Revision:
https://reviews.llvm.org/D37251
Change-Id: Ic2cf1d76598110401168326d411128ae2580a604
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316288 91177308-0d34-0410-b5e6-96231b3b80d8
This introduces a new operand type to encode the whether the index register should be XMM/YMM/ZMM. And new code to fixup the results created by readSIB.
This has the nice effect of removing a bunch of code that hard coded the name of every GATHER and SCATTER instruction to map the index type.
This fixes PR32807.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@316273 91177308-0d34-0410-b5e6-96231b3b80d8
