The compiler was generating code that ends up ignoring a multiple
latency dependence between two instructions by scheduling the
intructions in back-to-back packets. 

The packetizer needs to end a packet if the latency of the current
current insruction and the source in the previous packet is
greater than 1 cycle. This case occurs when there is still room in
the current packet, but scheduling the instruction causes a stall.
Instead, the packetizer should start a new packet. Also, if the
current packet already contains a stall, then it is okay to add
another instruction to the packet that also causes a stall. This
occurs when there are no instructions that can be scheduled in
between the producer and consumer instructions.

This patch changes the latency for loads to 2 cycles from 3 cycles.
This change refects that a load only needs to be separated by
one extra packet to eliminate the stall.

Patch by Ikhlas Ajbar.

llvm-svn: 301954

llvm-svn: 301953

Patch by Colin LeMahieu.

llvm-svn: 301952

Patch by Colin LeMahieu.

llvm-svn: 301951

Just let TTI's cost do this instead of arbitrarily restricting
this.

llvm-svn: 301950

Patch by Colin LeMahieu.

llvm-svn: 301949

With the forthcoming codeview::StringTable which a pdb::StringTable
would hold an instance of as one member, this ambiguity becomes
confusing.  Rename to PDBStringTable to avoid this.

llvm-svn: 301948

Patch by Colin LeMahieu.

llvm-svn: 301947

Change location number to point to conflicting branch instruction.

Patch by Colin LeMahieu.

llvm-svn: 301946

Patch by Colin LeMahieu.

llvm-svn: 301945

This reverts commit c08155afc5d3230792da2ad30a046a8617735a73.

This is causing undefined symbol errors with some of the constants.

llvm-svn: 301944

Patch by Colin LeMahieu.

llvm-svn: 301943

Instead of defining a custom iterator class, just use a function with a
callback, which is much easier to understand and less error prone.

Differential Revision: https://reviews.llvm.org/D32470

llvm-svn: 301942

If all jobs complete successfully, use pool.close() instead of
pool.terminate() before waiting for the workers. Zach Turner reported
that he was getting "access denied" exceptions from pool.terminate().

Make the workers abort immediately without printing to stderr when they
are interrupted.

Finally, catch exceptions when attempting to remove our temporary
testing directory. On abnormal exit, there can often be open handles
that haven't been cleaned up yet.

llvm-svn: 301941

Differential Revision: http://reviews.llvm.org/D32728

llvm-svn: 301940

  TLSDESC_ADD_LO12 relocations
Rearrange ordering in AArch64.def to follow relocation encoding
Fix name:
  R_AARCH64_P32_LD64_GOT_LO12_NC => R_AARCH64_P32_LD32_GOT_LO12_NC
Add support for several "TLS", "TLSGD", and "TLSLD" relocations for
  ILP32
Fix return values from isNonILP32reloc
Add implementations for
  R_AARCH64_ADR_PREL_PG_HI21_NC, R_AARCH64_P32_LD32_GOT_LO12_NC,
  R_AARCH64_P32_TLSIE_LD32_GOTTPREL_LO12_NC,
  R_AARCH64_P32_TLSDESC_LD32_LO12, R_AARCH64_LD64_GOT_LO12_NC,
  *TLSLD_LDST128_DTPREL_LO12, *TLSLD_LDST128_DTPREL_LO12_NC,
  *TLSLE_LDST128_TPREL_LO12, *TLSLE_LDST128_TPREL_LO12_NC
Modify error messages to give name of equivalent relocation in the
  ABI not being used, along with better checking for non-existent
  requested relocations.
Added assembler support for "pg_hi21_nc"
Relocation definitions added without implementations:
  R_AARCH64_P32_TLSDESC_ADR_PREL21, R_AARCH64_P32_TLSGD_ADR_PREL21,
  R_AARCH64_P32_TLSGD_ADD_LO12_NC, R_AARCH64_P32_TLSLD_ADR_PREL21, 
  R_AARCH64_P32_TLSLD_ADR_PAGE21, R_AARCH64_P32_TLSLD_ADD_LO12_NC,
  R_AARCH64_P32_TLSLD_LD_PREL19, R_AARCH64_P32_TLSDESC_LD_PREL19,
  R_AARCH64_P32_TLSGD_ADR_PAGE21, R_AARCH64_P32_TLS_DTPREL,
  R_AARCH64_P32_TLS_DTPMOD, R_AARCH64_P32_TLS_TPREL,
  R_AARCH64_P32_TLSDESC
Fix encoding:
  R_AARCH64_P32_TLSDESC_ADR_PAGE21

Reviewers: Peter Smith

Patch by: Joel Jones (jjones@cavium.com)

Differential Revision: https://reviews.llvm.org/D32072

llvm-svn: 301939

Avoid analyzing functions multiple times. This allows
asserting that each function is only analyzed once.

llvm-svn: 301938

llvm-svn: 301937

Previously we wrote line information and file checksum
information, but we did not write information about inlinee
lines and functions.  This patch adds support for that.

llvm-svn: 301936

http://lab.llvm.org:8011/builders/clang-ppc64le-linux/builds/5941

llvm-svn: 301935

It can confuse bots collecting errors.

llvm-svn: 301934

llvm-svn: 301932

Strip on OpenBSD does not correctly handle an empty .eh_frame section
and produces broken binaries in that case. Currently lld creates such
an empty .eh_frame section, despite the fact that the OpenBSD crtend.o
explicitly inserts a terminator. The Linux LSB "standard":

https://refspecs.linuxfoundation.org/LSB_5.0.0/LSB-Core-generic/LSB-Core-generic/ehframechpt.html#EHFRAME

explicitly says that

    The .eh_frame section shall contain 1 or more Call Frame Information (CFI) records.

This diff includes a test that specifically tests the issue I'm seeing
on OpenBSD.

Patch by Mark Kettenis!

llvm-svn: 301931

Reviewers: arsenm, nhaehnle

Subscribers: mehdi_amini, kzhuravl, wdng, yaxunl, dstuttard, tpr, llvm-commits, t-tye

Differential Revision: https://reviews.llvm.org/D32644

llvm-svn: 301930

This was originally checked in here:
https://reviews.llvm.org/rL301923

And reverted here:
https://reviews.llvm.org/rL301924

Because there's a clang test that would fail after this. I fixed/removed the
offending CHECK lines in:
https://reviews.llvm.org/rL301928

So let's try this again. Original commit message:

This is the fold that causes the infinite loop in BoringSSL
(https://github.com/google/boringssl/blob/master/crypto/cipher/e_rc2.c)
when we fix instcombine demanded bits to prefer 'not' ops as in https://reviews.llvm.org/D32255.

There are 2 or 3 problems with dyn_castNotVal, and I don't think we can
reinstate https://reviews.llvm.org/D32255 until dyn_castNotVal is completely eliminated.

1. As shown here, it transforms 'not' into random xor. This transform is harmful to SCEV and codegen because 'not' can often be folded while random xor cannot.
2. It does not transform vector constants. This is actually a good thing, but if you don't believe the above argument, then we shouldn't have excluded vectors.
3. It tries to avoid transforming not(not(X)). That's nice, but it doesn't match the greedy nature of instcombine. If we DeMorganize a pattern that has an extra 'not' in it: ~(~(~X) & Y) --> (~X | ~Y)

  That's just another case of DeMorgan, so we should trust that we'll fold that pattern too: (~X | ~ Y) --> ~(X & Y)

Differential Revision: https://reviews.llvm.org/D32665

llvm-svn: 301929

Don't test the optimizer as part of front-end verification.

llvm-svn: 301928

The fast reset for large memory regions is not working
only on windows. So enable it for Go/linux/darwin/freebsd.

See https://github.com/golang/go/issues/20139
for background and motivation.

Based on idea by Josh Bleecher Snyder.

llvm-svn: 301927

Summary:
TSan's Android `__get_tls()` and `TLS_SLOT_TSAN` can be used by other sanitizers as well (see D32649), this change moves them to sanitizer_common.
I picked sanitizer_linux.h as their new home.
In the process, add the 32-bit versions for ARM, i386 & MIPS.

Can the address of `__get_tls()[TLS_SLOT_TSAN]` change in between the calls?
I am not sure if there is a need to repeat the construct as opposed to using a variable. So I left things as they were.

Testing on my side was restricted to a successful cross-compilation.

Reviewers: dvyukov, kubamracek

Reviewed By: dvyukov

Subscribers: aemerson, rengolin, srhines, dberris, arichardson, llvm-commits

Differential Revision: https://reviews.llvm.org/D32705

llvm-svn: 301926

llvm-svn: 301925

There's a clang test that is wrongly using -O1 and failing after this commit.

llvm-svn: 301924

This is the fold that causes the infinite loop in BoringSSL 
(https://github.com/google/boringssl/blob/master/crypto/cipher/e_rc2.c) 
when we fix instcombine demanded bits to prefer 'not' ops as in D32255.

There are 2 or 3 problems with dyn_castNotVal, and I don't think we can 
reinstate D32255 until dyn_castNotVal is completely eliminated.
1. As shown here, it transforms 'not' into random xor. This transform is 
   harmful to SCEV and codegen because 'not' can often be folded while 
   random xor cannot.
2. It does not transform vector constants. This is actually a good thing, 
   but if you don't believe the above argument, then we shouldn't have 
   excluded vectors.
3. It tries to avoid transforming not(not(X)). That's nice, but it doesn't
   match the greedy nature of instcombine. If we DeMorganize a pattern 
   that has an extra 'not' in it:
   ~(~(~X) & Y) --> (~X | ~Y)

   That's just another case of DeMorgan, so we should trust that we'll fold
   that pattern too:
   (~X | ~ Y) --> ~(X & Y)

Differential Revision: https://reviews.llvm.org/D32665

llvm-svn: 301923

Summary: This is a common pattern that arise when legalizing large integers operations. Only do it when Y + 1 cannot overflow as this would change the carry behavior of uaddo .

Reviewers: jyknight, nemanjai, mkuper, spatel, RKSimon, zvi, bkramer

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D32687

llvm-svn: 301922

This is motivated by https://reviews.llvm.org/D32488 where I am trying
to add printing of the section type for incompatible sections to LLD
error messages. This patch allows us to use the same code in
llvm-readobj and LLD instead of duplicating the function inside LLD.

Patch by Alexander Richardson!

llvm-svn: 301921

Document the 'code' data type, and that value{15-17} is different to 
value{17-15}.

Patch by @chenwj (Wei-Ren Chen).

Differential Revision: https://reviews.llvm.org/D32117

llvm-svn: 301920

Summary: Common pattern when legalizing large integers operations. Similar to D32687, when the carry isn't used.

Reviewers: jyknight, nemanjai, mkuper, spatel, RKSimon, zvi, bkramer

Differential Revision: https://reviews.llvm.org/D32738

llvm-svn: 301919

llvm-svn: 301918

Summary:
If we have symbol information in a separate file, we need to be very
careful about presenting a unified section view of module to the rest of
the debugger. ObjectFileELF had code to handle that, but it was being
overly cautious -- the section->GetFileSize()!=0 meant that the
unification would fail for sections which do not occupy any space in the
object file (e.g., .bss). In my case, that manifested itself as not
being able to display the values of .bss variables properly as the
section associated with the variable did not have it's load address set
(because it was not present in the unified section list).

I test this by making sure the unified section list and the variables
refer to the same section.

Reviewers: eugene, zturner

Subscribers: tberghammer, lldb-commits, mgorny

Differential Revision: https://reviews.llvm.org/D32434

llvm-svn: 301917

It leads to clang crashing, e.g. on this short code fragment (added to
test/SemaCXX/warn-thread-safety-parsing.cpp):

  class SomeClass {
  public:
    void foo() {
      auto l = [this] { auto l = [] EXCLUSIVE_LOCKS_REQUIRED(mu_) {}; };
    }
    Mutex mu_;
  };

llvm-svn: 301916

llvm-svn: 301915

Reviewers: ioeric

Reviewed By: ioeric

Subscribers: cfe-commits

Differential Revision: https://reviews.llvm.org/D32741

llvm-svn: 301914