Patch by Arrowdodger.

llvm-svn: 140064

llvm-svn: 140050

llvm-svn: 140048

llvm-svn: 140047

yet legal according to comments in LegalizeDAG.cpp:227. 

Memcpy nodes created for copying byval arguments are inserted before
CALLSEQ_START.

The two failing tests reported in PR10876 pass after applying this patch.  

llvm-svn: 140046

llvm-svn: 140045

llvm-svn: 140044

llvm-svn: 140043

llvm-svn: 140042

Specify an additional fixed bit in the Thumb2 SSAT encoding to prevent the decoder from emitting gibberish for this invalid encoding.

llvm-svn: 140041

Fix a typo in the bitcode reader in the handling of atomic stores.  Reported by David Meyer on llvmdev.

llvm-svn: 140040

For example, 'ldrb r9, [sp]!' is odd, but valid.

llvm-svn: 140035

Report missing template arguments more helpfully by supplying the name
of the missing argument in the error message.

llvm-svn: 140034

Floating point stack inline asm works.

llvm-svn: 140033

Handle STRT (and friends) like LDRT (and friends) for decoding purposes.  Port over additional encoding tests to decoding tests.

llvm-svn: 140032

llvm-svn: 140029

Add a MachO-specific "mode" to llvm-objdump, that, if enabled, gathers additional information that are only available on MachO.

- It can take FunctionStarts from a binary to find entry points more accurately.
- Symbol offsets in executables are correct now.

llvm-svn: 140028

llvm-svn: 140027

llvm-svn: 140026

llvm-svn: 140025

llvm-svn: 140024

llvm-svn: 140012

llvm-svn: 140011

llvm-svn: 140010

llvm-svn: 140003

llvm-svn: 140001

llvm-svn: 140000

Note that this example doesn't work anymore!

llvm-svn: 139999

llvm-svn: 139998

dag-combine optimization to implement the ext-load efficiently (using shuffles).

For example the type <4 x i8> is stored in memory as i32, but it needs to
find its way into a <4 x i32> register. Previously we scalarized the memory
access, now we use shuffles.

llvm-svn: 139995

llvm-svn: 139994

llvm-svn: 139993

llvm-svn: 139992

maxps and maxpd).  This broke the sse41-blend.ll testcase by causing
maxpd to be produced rather than a cmp+blend pair, which is the reason
I tweaked it.  Gives a small speedup on doduc with dragonegg when the
GCC vectorizer is used.

llvm-svn: 139986

mayLoad = 1

llvm-svn: 139973

llvm-svn: 139972

llvm-svn: 139970

llvm-svn: 139969

Add ImmutableMapRef and ImmutableSetRef, which consolidate Immutable[Map,Set] and its Factory.  This may eventually replace Immtuable[Map,Set].

llvm-svn: 139967

llvm-svn: 139966