Hal Finkel [Fri, 4 Oct 2013 23:41:05 +0000 (23:41 +0000)]
UpdatePHINodes in BasicBlockUtils should not crash on duplicate predecessors
UpdatePHINodes has an optimization to reuse an existing PHI node, where it
first deletes all of its entries and then replaces them. Unfortunately, in the
case where we had duplicate predecessors (which are allowed so long as the
associated PHI entries have the same value), the loop removing the existing PHI
entries from the to-be-reused PHI would assert (if that PHI was not the one
which had the duplicates).
but the implementation does not handle indexed loads (pre/post inc.), but did
not specifically ignore them either (unlike for extending loads, which it
already ignored), causing an assert when the transformation was applied to an
indexed load. This is the minimal fix for correctness (causing the
transformation to be skipped for indexed loads).
Jack Carter [Fri, 4 Oct 2013 21:26:15 +0000 (21:26 +0000)]
[MC][AsmParser] Hook for post assembly file processing
This patch handles LLVM standalone assembler (llvm-mc) ELF flag setting based on input file
directive processing.
Mips assembly requires processing inline directives that directly and
indirectly affect the output ELF header flags. This patch handles one
".abicalls".
To process these directives we are following the model the code generator
uses by storing state in a container as we go through processing and when
we detect the end of input file processing, AsmParser is notified and we
update the ELF header flags through a MipsELFStreamer method with a call from
MCTargetAsmParser::emitEndOfAsmFile(MCStreamer &OutStreamer).
This patch will allow other targets the same functionality.
Make sure that we recursively vectorize <%conv0, %conv1, %conv2> and <%load0,
%load0, %load0>.
This makes it more likely to obtain vectorizable trees. We have to be careful
when we sort that we don't destroy 'good' existing ordering implied by source
order.
Temporarily revert r191792 as it is causing some LTO debug failures
on platforms with relocations in debug info and also temporarily
revert r191800 due to conflicts with the revert of r191792.
Matthias Braun [Fri, 4 Oct 2013 16:52:56 +0000 (16:52 +0000)]
ARM: optimizeSelect has to consider the previous register class
optimizeSelect folds (predicated) copy instructions, it must not ignore
the original register class of the operand when replacing the register
with the copies dest register.
Matthias Braun [Fri, 4 Oct 2013 16:52:54 +0000 (16:52 +0000)]
ARM: do not add a regmask for TAILJUMPs
The jump doesn't really kill the registers, the following call does but
we never get back anyway.
This avoids some verify-machineinstrs problems when TAILJUMPs are
if-converted.
Craig Topper [Fri, 4 Oct 2013 05:22:20 +0000 (05:22 +0000)]
Add OPC_CheckChildSame0-3 to the DAG isel matcher. This replaces sequences of MoveChild, CheckSame, MoveParent. Saves 846 bytes from the X86 DAG isel matcher, ~300 from ARM, ~840 from Hexagon.
David Blaikie [Fri, 4 Oct 2013 01:39:59 +0000 (01:39 +0000)]
DebugInfo: Fix ordering of members after r191928
In the case (shown in the attached test) where a member function
definition was emitted into debug info the following could occur:
1) build the debug info for the member function definition
2) in (1), build the debug info for the member function declaration
3) construct and add the member function declaration DIE
4) add it to its context
5) build its context (the type it is a member of)
6) construct the members and add them to the type
7) except don't add member functions because "getOrCreateSubprogram"
adds the function to its parent anyway
8) except we're only partway through building this subprogram
declaration so it hasn't been added yet - but we returned the partially
constructed DIE (since it's already in the MDNode->DIE mapping to avoid
infinitely recursing trying to create the member function DIE)
9) once the type is constructed, add the member function to it
10) now the members are out of order (the member function being defined
is listed as the last member, even though it was declared as the first)
To avoid this, construct the context of the subprogram DIE before we
query to see if it exists. That way we never end up creating it before
creating its context and ending up in this situation.
Alternatively, the type construction that visits/builds all the members
could call something like getOrCreateSubprogram, but that doesn't ever
do the "add to context" step. Then the type building code would always
be responsible for adding members (and the subprogram "addToContextDIE"
would no-op because the context building would have added the subprogram
declaration to the type/context DIE already).
(the test cases updated were overly-sensitive to offsets or abbreviation
numbers. We don't have a nice way to make these tests more robust as yet
- multiline FileCheck matches would be required)
Richard Mitton [Thu, 3 Oct 2013 22:07:08 +0000 (22:07 +0000)]
Fixed a bug with section names containing special characters.
Changed the dwarf aranges code to not use getLabelEndName, as it turns out it's not reliable to call that given user-defined section names. Section names can have characters in that aren't representable as symbol names.
The dwarf-aranges test case has been updated to include a special character, to check this.
Owen Anderson [Thu, 3 Oct 2013 21:08:05 +0000 (21:08 +0000)]
Pull fptrunc's upwards through selects when one of the select's selectands was a constant. This has a number of benefits, including producing small immediates (easier to materialize, smaller constant pools) as well as being more likely to allow the fptrunc to fuse with a preceding instruction (truncating selects are unusual).
David Blaikie [Thu, 3 Oct 2013 20:07:20 +0000 (20:07 +0000)]
DebugInfo: Avoid redundantly adding child DIEs to parents.
DIE::addChild had a shortcircuit that silently no-op'd when a child was
readded to the same parent. This hid some quirky/redundant code in
DwarfDebug/CompileUnit. By removing that functionality and replacing it
with an assert I was able to find and cleanup those cases, mostly
centering around adding members to types in various circumstances.
1) The original oddity I noticed while working on type units (which
actually was helping me in the short term, by accident) was the
addToContextOwner call in constructTypeDIE. This call was completely
bogus (why was it only done for non-virtual types? what relevance does
that have at all) and redundant with the more uniform addToContextOwner
made in getOrCreateTypeDIE.
2) If a member function definition was visited (createSubprogramDIE), it
would attempt to build the member function declaration. The declaration
DIE would then be added to its context, but in building the context (the
type for which this function is a member) the members of the type would
be added to the type automatically, so by the time the context was
constructed, the member function was already associated with it.
3) The same as (2) but without the member function being constructed
first. Whenever a type was constructed, the members would be created and
member functions would be created by getOrCreateSubprogramDIE - this
would lead to the subprogram being added to the (incomplete) type
already, then the general member-construction code would add it again.
Optimize linkonce_odr unnamed_addr functions during LTO.
Generalize the API so we can distinguish symbols that are needed just for a DSO
symbol table from those that are used from some native .o.
The symbols that are only wanted for the dso symbol table can be dropped if
llvm can prove every other dso has a copy (linkonce_odr) and the address is not
important (unnamed_addr).
Make sure we emit a section for pubnames even if that section is
going to be empty. This is particularly important for the gnu
pubnames case since we're emitting a relocation to the section.
Benjamin Kramer [Thu, 3 Oct 2013 13:24:02 +0000 (13:24 +0000)]
CaptureTracking: Plug a loophole in the "too many uses" heuristic.
The heuristic was added to avoid spending too much compile time A specially
crafted test case (PR17461, PR16474) with many uses on a select or bitcast
instruction can still trigger the slow case. Add a check for that case.
This only affects compile time, don't have a good way to test it.
Alexey Samsonov [Thu, 3 Oct 2013 08:54:43 +0000 (08:54 +0000)]
Remove wild .debug_aranges entries generated from unimportant labels
r191052 added emitting .debug_aranges to Clang, but this
functionality is broken: it uses all MC labels added in DWARF Asm
printer, including the labels for build relocations between
different DWARF sections, like .Lsection_line or .Ldebug_loc0.
As a result, if any DIE .debug_info would contain "DW_AT_location=0x123"
attribute, .debug_aranges would also contain a range starting from 0x123,
breaking tools that rely on this section.
This patch fixes this by using only MC labels that corresponds to the
addresses in the user program.
[llvm-c][Disassembler] When printing latency information, skip scheduling
classes that are marked as Variant as those require an MI to pass to
SubTargetInfo::resolveSchedClass.
Matt Arsenault [Wed, 2 Oct 2013 22:38:17 +0000 (22:38 +0000)]
Don't use runtime bounds check between address spaces.
Don't vectorize with a runtime check if it requires a
comparison between pointers with different address spaces.
The values can't be assumed to be directly comparable.
Previously it would create an illegal bitcast.
[llvm-c][Disassembler] Add an option to print latency information in
disassembled output alongside the instructions.
E.g., on a vector shuffle operation with a memory operand, disassembled
outputs are:
* Without the option:
vpshufd $-0x79, (%rsp), %xmm0
* With the option:
vpshufd $-0x79, (%rsp), %xmm0 ## Latency: 5
The printed latency is extracted from the schedule model available in the
disassembler context. Thus, this option has no effect if there is not a
scheduling model for the target.
This boils down to one may need to specify the CPU string, so that this
option could have an effect.
Tom Stellard [Wed, 2 Oct 2013 17:04:59 +0000 (17:04 +0000)]
StructurizeCFG: Add dependency on LowerSwitch pass
Switch instructions were crashing the StructurizeCFG pass, and it's
probably easier anyway if we don't need to handle them in this pass.
Reviewed-by: Christian König <christian.koenig@amd.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@191841 91177308-0d34-0410-b5e6-96231b3b80d8
Remove the very substantial, largely unmaintained legacy PGO
infrastructure.
This was essentially work toward PGO based on a design that had several
flaws, partially dating from a time when LLVM had a different
architecture, and with an effort to modernize it abandoned without being
completed. Since then, it has bitrotted for several years further. The
result is nearly unusable, and isn't helping any of the modern PGO
efforts. Instead, it is getting in the way, adding confusion about PGO
in LLVM and distracting everyone with maintenance on essentially dead
code. Removing it paves the way for modern efforts around PGO.
Among other effects, this removes the last of the runtime libraries from
LLVM. Those are being developed in the separate 'compiler-rt' project
now, with somewhat different licensing specifically more approriate for
runtimes.
Don't layout items in a list in columns. That requires changing every
line just to add or remove a single element. What I wouldn't give to
have clang-format here an be able to format this more densely without
caring...
Re-group and sort the entries while here to make the whole thing more
clear.
projects / llvm / log