Vince Weaver
e81cc233a6
X86: Remove double-cast in Cvtf2i micro-op
...
This double cast led to rounding errors which caused
some benchmarks to get the wrong values, most notably lucas
which failed spectacularly due to CVTTSD2SI returning an
off-by-one value. equake was also broken.
2009-11-10 11:18:23 -05:00
Vince Weaver
5cf2e7ccf0
X86: Fix problem with movhps instruction
...
This problem is like the one fixed with movhpd a few weeks ago.
A +8 displacement is used to access memory when there should
be none.
This fix is needed for the perlbmk spec2k benchmark to run.
2009-11-04 13:22:15 -05:00
Vince Weaver
a12557439b
X86: Add support for x86 psrldq and pslldq instructions
...
These are complicated instructions and the micro-code might be suboptimal.
This has been tested with some small sample programs (attached)
The psrldq instruction is needed by various spec2k programs.
2009-10-30 12:49:37 -04:00
Vince Weaver
5873ec2238
X86: Implement movd_Vo_Edp on X86
...
This patch implements the movd_Vo_Edp series of instructions.
It addresses various concerns by Gabe Black about which file the
instruction belonged in, as well as supporting REX prefixed
instructions properly.
This instruction is needed for some of the spec2k benchmarks, most
notably bzip2.
2009-10-30 15:52:33 -04:00
Vince Weaver
b2067840a6
X86: Implement the X86 sse2 haddpd instruction
...
This patch implements the haddpd instruction.
It fixes the problem in the previous version (pointed out by Gabe Black)
where an incorrect result would happen if you issue the instruction
with the same argument twice, i.e. "haddpd %xmm0,%xmm0"
This instruction is used by many spec2k benchmarks.
2009-10-30 14:19:06 -04:00
Gabe Black
f9624e49f6
X86: Replace "DISPLACEMENT" with disp in movhpd.
2009-10-27 23:50:25 -07:00
Vince Weaver
87b97f28bd
Fix problem with the x86 sse movhpd instruction.
...
The movhpd instruction was writing to the wrong memory offset.
2009-10-27 14:11:06 -04:00
Vince Weaver
14691148cd
Implement X86 sse2 movdqu and movdqa instructions
...
The movdqa instruction should enforce 16-byte alignment.
This implementation does not do that.
These instructions are needed for most of x86_64 spec2k to run.
2009-10-21 13:40:43 -04:00
Gabe Black
c876a781a5
X86: Sign extend the immediate of wripi like the register version.
2009-09-16 19:29:51 -07:00
Gabe Black
7a0ef6c36f
X86: Make the imm8 member of immediate microops really 8 bits consistently.
2009-09-16 19:28:57 -07:00
Gabe Black
239f1dea31
X86: Fix checking the NT bit during an IRET.
2009-09-16 19:28:30 -07:00
Gabe Black
eec6bfaa9d
X86: Fix setting the busy bit in the task descriptor in LTR.
2009-09-16 19:28:01 -07:00
Gabe Black
e251b42c59
Merge with head.
2009-08-23 14:19:14 -07:00
Gabe Black
d0d597004f
X86: Preserve the NO_ACCESS flag when giving CDA a specialized interface.
2009-08-23 14:16:58 -07:00
Nathan Binkert
890be77362
X86: fix some simple compile issues
...
static should not be used for constants that are not inside a class definition.
2009-08-21 09:10:25 -07:00
Gabe Black
e8c0ca5cd1
X86: Fix the decoding for and fill out FST and FSTP.
2009-08-20 00:42:14 -07:00
Gabe Black
843d064668
X86: Add microassembler symbols for floating point stack register operands.
2009-08-20 00:41:27 -07:00
Gabe Black
ed088ed15e
X86: Decode the immediate byte opcode extension for 3dNow! instructions.
2009-08-18 00:52:47 -07:00
Gabe Black
fd45c04cad
X86: Decode three byte opcodes.
2009-08-18 00:52:45 -07:00
Gabe Black
9df74ab401
X86: Double check the two byte portion of the decoder and fix bugs/clean up.
2009-08-17 20:25:15 -07:00
Gabe Black
92867cac95
X86: Implement MOVNTI.
2009-08-17 20:25:15 -07:00
Gabe Black
56d87cdc08
X86: Implement MOVQ2DQ.
2009-08-17 20:25:14 -07:00
Gabe Black
317a9ac6d1
X86: Implement MOVDQ2Q.
2009-08-17 20:25:14 -07:00
Gabe Black
1606663aa9
X86: Implement the media instructions that convert fp values to ints.
2009-08-17 20:25:14 -07:00
Gabe Black
e3ef432a55
X86: Implement a microop for converting fp values to ints.
2009-08-17 20:25:14 -07:00
Gabe Black
123ea3b229
X86: Implement the instructions that compare fp values and write a mask as a result.
2009-08-17 20:25:14 -07:00
Gabe Black
288f428632
X86: Implement a microop that compares fp values and writes a mask as a result.
2009-08-17 20:25:14 -07:00
Gabe Black
87ad677209
X86: Implement the instructions that compare fp values and write to rflags.
2009-08-17 20:25:14 -07:00
Gabe Black
2c9ee52c37
X86: Implement a microop that compares fp values and writes to rflags.
2009-08-17 20:25:14 -07:00
Gabe Black
7d4db7266e
X86: Implement MOVSS.
2009-08-17 20:25:14 -07:00
Gabe Black
179fd4e536
X86: Implement LDMXCSR.
2009-08-17 20:25:14 -07:00
Gabe Black
b315c3effc
X86: Implement STMXCSR.
2009-08-17 20:25:13 -07:00
Gabe Black
7fbd9e1296
X86: Implement the shuffle media instructions.
2009-08-17 20:25:13 -07:00
Gabe Black
1fed0161d9
X86: Implement a shuffle media microop.
2009-08-17 20:25:13 -07:00
Gabe Black
d9970f139a
X86: Implement the mask move instructions.
2009-08-17 20:22:56 -07:00
Gabe Black
75528a497c
X86: Implement a mask move microop.
2009-08-17 20:22:56 -07:00
Gabe Black
90dc1abd0b
X86: Implement the instructions that move sign bits.
2009-08-17 20:22:56 -07:00
Gabe Black
90786e43fc
X86: Implement a microop that moves sign bits.
2009-08-17 20:22:56 -07:00
Gabe Black
4c23e631f2
X86: Implement the insert/extract instructions.
2009-08-17 20:22:56 -07:00
Gabe Black
c0e850c77a
X86: Fix a bug in the decoder where the insert/extract instructions go.
2009-08-17 20:22:56 -07:00
Gabe Black
965e546df3
X86: Extend mov2int and mov2fp so they can support insert and extract instructions.
2009-08-17 20:22:56 -07:00
Gabe Black
2beab367d7
X86: Implement the media average instructions.
2009-08-17 20:15:16 -07:00
Gabe Black
f6b12bfa8d
X86: Implement a media average microop.
2009-08-17 20:15:16 -07:00
Gabe Black
45bae0c70f
X86: Implement the multiply and add instructions.
2009-08-17 20:15:16 -07:00
Gabe Black
200fed31de
X86: Let the integer multiply microop use every other possible source value.
2009-08-17 20:15:16 -07:00
Gabe Black
cf2fc2613d
X86: Implement the media shifts that operate on 64 bits or less at a time.
2009-08-17 20:15:16 -07:00
Gabe Black
c8a0cf5df7
X86: Implement the media shift microops. These don't handle full 128 bit wide shifts.
2009-08-17 20:15:16 -07:00
Gabe Black
dabbc7d9d3
X86: Implement the sum of absolute differences instructions.
2009-08-17 20:15:16 -07:00
Gabe Black
470dcef229
X86: Implement a "sum of absolute differences" microop.
2009-08-17 20:15:16 -07:00
Gabe Black
31d29ee3f8
X86: Implement the media integer subtract instructions.
2009-08-17 20:15:15 -07:00
