load source operand is used by multiple nodes. The v2i64 broadcast was emulated
by shuffling the two lower i32 elements to the upper two.
We had a bug in the immediate used for the broadcast.
Replacing 0 to 0x44.
0x44 means [01|00|01|00] which corresponds to the correct lane.
Patch by Michael Kuperstein.
llvm-svn: 160430
Print the high order register of a double word register operand.
In 32 bit mode, a 64 bit double word integer will be represented
by 2 32 bit registers. This modifier causes the high order register
to be used in the asm expression. It is useful if you are using
doubles in assembler and continue to control register to variable
relationships.
This patch also fixes a related bug in a previous patch:
case 'D': // Second part of a double word register operand
case 'L': // Low order register of a double word register operand
case 'M': // High order register of a double word register operand
I got 'D' and 'M' confused. The second part of a double word operand
will only match 'M' for one of the endianesses. I had 'L' and 'D'
be the opposite twins when 'L' and 'M' are.
llvm-svn: 160429
Fixes PR13371: indvars pass incorrectly substitutes 'undef' values.
I do not like this fix. It's needed until/unless the meaning of undef
changes. It attempts to be complete according to the IR spec, but I
don't have much confidence in the implementation given the difficulty
testing undefined behavior. Worse, this invalidates some of my
hard-fought work on indvars and LSR to optimize pointer induction
variables. It results benchmark regressions, which I'll track
internally. On x86_64 no LTO I see:
-3% huffbench
-3% 400.perlbench
-8% fhourstones
My only suggestion for recovering is to change the meaning of
undef. If we could trust an arbitrary instruction to produce a some
real value that can be manipulated (e.g. incremented) according to
non-undef rules, then this case could be easily handled with SCEV.
llvm-svn: 160421
be defined as deleted, take cv-qualifiers on class members into account when
looking up the copy or move constructor or assignment operator which will be
used for them.
llvm-svn: 160418
and a function template instantiation, if there's a parameter pack in the
declaration and one at the same place in the instantiation, don't assume that
the pack wasn't expanded -- it may have expanded to nothing. Instead, go ahead
and check whether the parameter pack was expandable. We can do this as a
side-effect of the work we'd need to do anyway, to find how many parameters
were produced.
llvm-svn: 160416
[temp.deduct.call]p4 under Objective-C++ ARC, make sure to adjust the
qualifiers to introduce the implicit strong lifetime when
needed. Fixes <rdar://problem/11825671>.
llvm-svn: 160412
intrinsics. The second instruction(s) to be handled are the vector versions
of count set bits (ctpop).
The changes here are to clang so that it generates a target independent
vector ctpop when it sees an ARM dependent vector bits set count. The changes
in llvm are to match the target independent vector ctpop and in
VMCore/AutoUpgrade.cpp to update any existing bc files containing ARM
dependent vector pop counts with target-independent ctpops. There are also
changes to an existing test case in llvm for ARM vector count instructions and
to a test for the bitcode upgrade.
<rdar://problem/11892519>
There is deliberately no test for the change to clang, as so far as I know, no
consensus has been reached regarding how to test neon instructions in clang;
q.v. <rdar://problem/8762292>
llvm-svn: 160410
intrinsics. The second instruction(s) to be handled are the vector versions
of count set bits (ctpop).
The changes here are to clang so that it generates a target independent
vector ctpop when it sees an ARM dependent vector bits set count. The changes
in llvm are to match the target independent vector ctpop and in
VMCore/AutoUpgrade.cpp to update any existing bc files containing ARM
dependent vector pop counts with target-independent ctpops. There are also
changes to an existing test case in llvm for ARM vector count instructions and
to a test for the bitcode upgrade.
<rdar://problem/11892519>
There is deliberately no test for the change to clang, as so far as I know, no
consensus has been reached regarding how to test neon instructions in clang;
q.v. <rdar://problem/8762292>
llvm-svn: 160409
given declaration.
It is based on the observation that during parsing the comment that should be
attached to the decl is usually among the last two documentation comments
parsed.
llvm-svn: 160400
Remove assertions and turn what used the be the assertion into a logged error with instructions on what to attach to a radar so we can track down why this is happening.
llvm-svn: 160392
A standalone pattern defined in a multiclass expansion should handle
null_frag references just like patterns on instructions. Follow-up to
r160333.
llvm-svn: 160384
Checks against nil often appear as guards in macros, and comparing
Objective-C literals to nil has well-defined behavior (if tautological).
On OS X, 'nil' has not been typed as 'id' since 10.6 (possibly earlier),
so the warning was already not firing, but other runtimes continue to use
((id)0) or some variant. This change accepts comparisons to any null pointer;
to keep it simple, it looks through all casts (not just casts to 'id').
PR13276
llvm-svn: 160379
Nadav Rotem