Use generic routine for 64-bit carry-less multiply.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This implements the AESDEC instruction.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This implements the AESENC instruction.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This implements the AESIMC instruction.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This implements the AESDECLAST instruction.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
This implements the AESENCLAST instruction.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
vzeroall: xmm_regs should be used instead of xmm_t0
vpermdq: bit 3 and 7 of imm should be considered
Signed-off-by: Xinyu Li <lixinyu20s@ict.ac.cn>
Message-Id: <20230510145222.586487-1-lixinyu20s@ict.ac.cn>
Cc: qemu-stable@nongnu.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
To avoid compilation errors when -Werror=maybe-uninitialized is used,
add a default case with g_assert_not_reached().
Otherwise with GCC 11.3.1 "cc (GCC) 11.3.1 20220421 (Red Hat 11.3.1-2)"
we get:
../target/i386/ops_sse.h: In function ‘helper_vpermdq_ymm’:
../target/i386/ops_sse.h:2495:13: error: ‘r3’ may be used
uninitialized in this function [-Werror=maybe-uninitialized]
2495 | d->Q(3) = r3;
| ~~~~~~~~^~~~
../target/i386/ops_sse.h:2494:13: error: ‘r2’ may be used
uninitialized in this function [-Werror=maybe-uninitialized]
2494 | d->Q(2) = r2;
| ~~~~~~~~^~~~
../target/i386/ops_sse.h:2493:13: error: ‘r1’ may be used
uninitialized in this function [-Werror=maybe-uninitialized]
2493 | d->Q(1) = r1;
| ~~~~~~~~^~~~
../target/i386/ops_sse.h:2492:13: error: ‘r0’ may be used
uninitialized in this function [-Werror=maybe-uninitialized]
2492 | d->Q(0) = r0;
| ~~~~~~~~^~~~
Signed-off-by: Eric Auger <eric.auger@redhat.com>
Message-Id: <20221222140158.1260748-1-eric.auger@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The only issue with FMA instructions is that there are _a lot_ of them (30
opcodes, each of which comes in up to 4 versions depending on VEX.W and
VEX.L; a total of 96 possibilities). However, they can be implement with
only 6 helpers, two for scalar operations and four for packed operations.
(Scalar versions do not do any merging; they only affect the bottom 32
or 64 bits of the output operand. Therefore, there is no separate XMM
and YMM of the scalar helpers).
First, we can reduce the number of helpers to one third by passing four
operands (one output and three inputs); the reordering of which operands
go to the multiply and which go to the add is done in emit.c.
Second, the different instructions also dispatch to the same softfloat
function, so the flags for float32_muladd and float64_muladd are passed
in the helper as int arguments, with a little extra complication to
handle FMADDSUB and FMSUBADD.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
F16C only consists of two instructions, which are a bit peculiar
nevertheless.
First, they access only the low half of an YMM or XMM register for the
packed-half operand; the exact size still depends on the VEX.L flag.
This is similar to the existing avx_movx flag, but not exactly because
avx_movx is hardcoded to affect operand 2. To this end I added a "ph"
format name; it's possible to reuse this approach for the VPMOVSX and
VPMOVZX instructions, though that would also require adding two more
formats for the low-quarter and low-eighth of an operand.
Second, VCVTPS2PH is somewhat weird because it *stores* the result of
the instruction into memory rather than loading it.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
VROUND, FSTCW and STMXCSR all have to perform the same conversion from
x86 rounding modes to softfloat constants. Since the ISA is consistent
on the meaning of the two-bit rounding modes, extract the common code
into a wrapper for set_float_rounding_mode.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
With all SSE (and AVX!) instructions now implemented in disas_insn_new,
it's possible to remove gen_sse, as well as the helpers for instructions
that now use gvec.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
These are mostly moves, and yet are a total pain. The main issue
is that:
1) some instructions are selected by mod==11 (register operand)
vs. mod=00/01/10 (memory operand)
2) stores to memory are two-operand operations, while the 3-register
and load-from-memory versions operate on the entire contents of the
destination; this makes it easier to separate the gen_* function for
the store case
3) it's inefficient to load into xmm_T0 only to move the value out
again, so the gen_* function for the load case is separated too
The manual also has various mistakes in the operands here, for example
the store case of MOVHPS operates on a 128-bit source (albeit discarding
the bottom 64 bits) and therefore should be Mq,Vdq rather than Mq,Vq.
Likewise for the destination and source of MOVHLPS.
VUNPCK?PS and VUNPCK?PD are the same as VUNPCK?DQ and VUNPCK?QDQ,
but encoded as prefixes rather than separate operands. The helpers
can be reused however.
For MOVSLDUP, MOVSHDUP and MOVDDUP I chose to reimplement them as
helpers. I named the helper for MOVDDUP "movdldup" in preparation
for possible future introduction of MOVDHDUP and to clarify the
similarity with MOVSLDUP.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
There are several special cases here:
1) extending moves have different widths for the helpers vs. for the
memory loads, and the width for memory loads depends on VEX.L too.
This is represented by X86_SPECIAL_AVXExtMov.
2) some instructions, such as variable-width shifts, select the vector element
size via REX.W.
3) VSIB instructions (VGATHERxPy, VPGATHERxy) are also part of this group,
and they have (among other things) two output operands.
3) the macros for 4-operand blends (which are under 0x0f 0x3a) have to be
extended to support 2-operand blends. The 2-operand variant actually
came a few years earlier, but it is clearer to implement them in the
opposite order.
X86_TYPE_WM, introduced earlier for unaligned loads, is reused for helpers
that accept a Reg* but have a M argument.
These three-byte opcodes also include AVX new instructions, for which
the helpers were originally implemented by Paul Brook <paul@nowt.org>.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The more complicated operations here are insertions and extractions.
Otherwise, there are just more entries than usual because the PS/PD/SS/SD
variations are encoded in the opcode rater than in the prefixes.
These three-byte opcodes also include AVX new instructions, whose
implementation in the helpers was originally done by Paul Brook
<paul@nowt.org>.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Three-byte opcodes from the 0F3Ah area all have an immediate byte which
is usually unsigned. Clarify in the helper code that it is unsigned;
the new decoder treats immediates as signed by default, and seeing
an intN_t in the prototype might give the wrong impression that one
can use decode->immediate directly.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The new implementation of SSE will cover AVX from the get go, because
all the work for the helper functions is already done. We just need to
build them.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The new implementation of SSE will cover AVX from the get go, so include
the 24 extra comparison operators that are only available with the VEX
prefix.
Based on a patch by Paul Brook <paul@nowt.org>.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Compared to Paul's implementation, the new decoder will use a different approach
to implement AVX's merging of dst with src1 on scalar operations. Adjust the
old SSE decoder to be compatible with new-style helpers.
The affected instructions are CVTSx2Sx, ROUNDSx, RSQRTSx, SQRTSx, RCPSx.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Compared to Paul's implementation, the new decoder will use a different approach
to implement AVX's merging of dst with src1 on scalar operations. Adjust the
helpers to provide this functionality.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add to the helpers all the operands that are needed to implement AVX.
Extracted from a patch by Paul Brook <paul@nowt.org>.
Message-Id: <20220424220204.2493824-26-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
INSERTQ is defined to not modify any bits in the lower 64 bits of the
destination, other than the ones being replaced with bits from the
source operand. QEMU instead is using unshifted bits from the source
for those bits.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
SSE4a instructions EXTRQ and INSERTQ have two bit index operands, that can be
immediates or taken from an XMM register. In both cases, the fields are
6-bit wide and the top two bits in the byte are ignored. translate.c is
doing that correctly for the immediate case, but not for the XMM case, so
fix it.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Make the AES vector helpers AVX ready
No functional changes to existing helpers
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-22-paul@nowt.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Make the pclmulqdq helper AVX ready
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-21-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Rewrite the blendv helpers so that they can easily be extended to support
the AVX encodings, which make all 4 arguments explicit.
No functional changes to the existing helpers
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-20-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Fixup various vector helpers that either trivially exten to 256 bit,
or don't have 256 bit variants.
No functional changes to existing helpers
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-19-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Perpare the horizontal atithmetic vector helpers for AVX
These currently use a dummy Reg typed variable to store the result then
assign the whole register. This will cause 128 bit operations to corrupt
the upper half of the register, so replace it with explicit temporaries
and element assignments.
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-18-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Make the dpps and dppd helpers AVX-ready
I can't see any obvious reason why dppd shouldn't work on 256 bit ymm
registers, but both AMD and Intel agree that it's xmm only.
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-17-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
AVX includes an additional set of comparison predicates, some of which
our softfloat implementation does not expose as separate functions.
Rewrite the helpers in terms of floatN_compare for future extensibility.
Signed-off-by: Paul Brook <paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20220424220204.2493824-24-paul@nowt.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Prepare the "easy" floating point vector helpers for AVX
No functional changes to existing helpers.
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-16-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
These helpers need to take special care to avoid overwriting source values
before the wole result has been calculated. Currently they use a dummy
Reg typed variable to store the result then assign the whole register.
This will cause 128 bit operations to corrupt the upper half of the register,
so replace it with explicit temporaries and element assignments.
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-14-paul@nowt.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
More preparatory work for AVX support in various integer vector helpers
No functional changes to existing helpers.
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-13-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Rewrite the "simple" vector integer helpers in preperation for AVX support.
While the current code is able to use the same prototype for unary
(a = F(b)) and binary (a = F(b, c)) operations, future changes will cause
them to diverge.
No functional changes to existing helpers
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-12-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Rewrite the vector shift helpers in preperation for AVX support (3 operand
form and 256 bit vectors).
For now keep the existing two operand interface.
No functional changes to existing helpers.
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-11-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
For AVX we're going to need both 128 bit (xmm) and 256 bit (ymm) variants of
floating point helpers. Add the register type suffix to the existing
*PS and *PD helpers (SS and SD variants are only valid on 128 bit vectors)
No functional changes.
Signed-off-by: Paul Brook <paul@nowt.org>
Message-Id: <20220424220204.2493824-15-paul@nowt.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The DPPS (Dot Product) instruction is defined to first sum pairs of
intermediate results, then sum those values to get the final result.
i.e. (A+B)+(C+D)
We incrementally sum the results, i.e. ((A+B)+C)+D, which can result
in incorrect rouding.
For consistency, also change the variable names to the ones used
in the Intel SDM and implement DPPD following the manual.
Based on a patch by Paul Brook <paul@nowt.org>.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The computation must not overwrite neither the destination
nor the source before the last element has been computed.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The abs1 function in ops_sse.h only works sorrectly when the result fits
in a signed int. This is fine most of the time because we're only dealing
with byte sized values.
However pcmp_elen helper function uses abs1 to calculate the absolute value
of a cpu register. This incorrectly truncates to 32 bits, and will give
the wrong anser for the most negative value.
Fix by open coding the saturation check before taking the absolute value.
Signed-off-by: Paul Brook <paul@nowt.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The i386 target consolidates all vector registers so that instead of
XMMReg, YMMReg and ZMMReg structs there is a single ZMMReg that can
fit all of SSE, AVX and AVX512.
When TCG copies data from and to the SSE registers, it uses the
full 64-byte width. This is not a correctness issue because TCG
never lets guest code see beyond the first 128 bits of the ZMM
registers, however it causes uninitialized stack memory to
make it to the CPU's migration stream.
Fix it by only copying the low 16 bytes of the ZMMReg union into
the destination register.
Reviewed-by: Peter Maydell <peter.maydell@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
There is no "version 2" of the "Lesser" General Public License.
It is either "GPL version 2.0" or "Lesser GPL version 2.1".
This patch replaces all occurrences of "Lesser GPL version 2" with
"Lesser GPL version 2.1" in comment section.
Signed-off-by: Chetan Pant <chetan4windows@gmail.com>
Message-Id: <20201023122801.19514-1-chetan4windows@gmail.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
The SSE instruction implementations all fail to raise the expected
IEEE floating-point exceptions because they do nothing to convert the
exception state from the softfloat machinery into the exception flags
in MXCSR.
Fix this by adding such conversions. Unlike for x87, emulated SSE
floating-point operations might be optimized using hardware floating
point on the host, and so a different approach is taken that is
compatible with such optimizations. The required invariant is that
all exceptions set in env->sse_status (other than "denormal operand",
for which the SSE semantics are different from those in the softfloat
code) are ones that are set in the MXCSR; the emulated MXCSR is
updated lazily when code reads MXCSR, while when code sets MXCSR, the
exceptions in env->sse_status are set accordingly.
A few instructions do not raise all the exceptions that would be
raised by the softfloat code, and those instructions are made to save
and restore the softfloat exception state accordingly.
Nothing is done about "denormal operand"; setting that (only for the
case when input denormals are *not* flushed to zero, the opposite of
the logic in the softfloat code for such an exception) will require
custom code for relevant instructions, or else architecture-specific
conditionals in the softfloat code for when to set such an exception
together with custom code for various SSE conversion and rounding
instructions that do not set that exception.
Nothing is done about trapping exceptions (for which there is minimal
and largely broken support in QEMU's emulation in the x87 case and no
support at all in the SSE case).
Signed-off-by: Joseph Myers <joseph@codesourcery.com>
Message-Id: <alpine.DEB.2.21.2006252358000.3832@digraph.polyomino.org.uk>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This corrects a bug introduced in my previous fix for SSE4.2 pcmpestri
/ pcmpestrm / pcmpistri / pcmpistrm substring search, commit
ae35eea7e4.
That commit fixed a bug that showed up in four GCC tests with one libc
implementation. The tests in question generate random inputs to the
intrinsics and compare results to a C implementation, but they only
test 1024 possible random inputs, and when the tests use the cases of
those instructions that work with word rather than byte inputs, it's
easy to have problematic cases that show up much less frequently than
that. Thus, testing with a different libc implementation, and so a
different random number generator, showed up a problem with the
previous patch.
When investigating the previous test failures, I found the description
of these instructions in the Intel manuals (starting from computing a
16x16 or 8x8 set of comparison results) confusing and hard to match up
with the more optimized implementation in QEMU, and referred to AMD
manuals which described the instructions in a different way. Those
AMD descriptions are very explicit that the whole of the string being
searched for must be found in the other operand, not running off the
end of that operand; they say "If the prototype and the SUT are equal
in length, the two strings must be identical for the comparison to be
TRUE.". However, that statement is incorrect.
In my previous commit message, I noted:
The operation in this case is a search for a string (argument d to
the helper) in another string (argument s to the helper); if a copy
of d at a particular position would run off the end of s, the
resulting output bit should be 0 whether or not the strings match in
the region where they overlap, but the QEMU implementation was
wrongly comparing only up to the point where s ends and counting it
as a match if an initial segment of d matched a terminal segment of
s. Here, "run off the end of s" means that some byte of d would
overlap some byte outside of s; thus, if d has zero length, it is
considered to match everywhere, including after the end of s.
The description "some byte of d would overlap some byte outside of s"
is accurate only when understood to refer to overlapping some byte
*within the 16-byte operand* but at or after the zero terminator; it
is valid to run over the end of s if the end of s is the end of the
16-byte operand. So the fix in the previous patch for the case of d
being empty was correct, but the other part of that patch was not
correct (as it never allowed partial matches even at the end of the
16-byte operand). Nor was the code before the previous patch correct
for the case of d nonempty, as it would always have allowed partial
matches at the end of s.
Fix with a partial revert of my previous change, combined with
inserting a check for the special case of s having maximum length to
determine where it is necessary to check for matches.
In the added test, test 1 is for the case of empty strings, which
failed before my 2017 patch, test 2 is for the bug introduced by my
2017 patch and test 3 deals with the case where a match of an initial
segment at the end of the string is not valid when the string ends
before the end of the 16-byte operand (that is, the case that would be
broken by a simple revert of the non-empty-string part of my 2017
patch).
Signed-off-by: Joseph Myers <joseph@codesourcery.com>
Message-Id: <alpine.DEB.2.21.2006121344290.9881@digraph.polyomino.org.uk>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Detected by asm test suite failures in dav1d
(https://code.videolan.org/videolan/dav1d). Can be reproduced by
`qemu-x86_64 -cpu core2duo ./tests/checkasm --test=mc_8bpc 1659890620`.
Signed-off-by: Janne Grunau <j@jannau.net>
Message-Id: <20200401225253.30745-1-j@jannau.net>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Give the previously unnamed enum a typedef name. Use it in the
prototypes of compare functions. Use it to hold the results
of the compare functions.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
The x86 architecture requires that all conversions from floating
point to integer which raise the 'invalid' exception (infinities of
both signs, NaN, and all values which don't fit in the destination
integer) return what the x86 spec calls the "indefinite integer
value", which is 0x8000_0000 for 32-bits or 0x8000_0000_0000_0000 for
64-bits. The softfloat functions return the more usual behaviour of
positive overflows returning the maximum value that fits in the
destination integer format and negative overflows returning the
minimum value that fits.
Wrap the softfloat functions in x86-specific versions which
detect the 'invalid' condition and return the indefinite integer.
Note that we don't use these wrappers for the 3DNow! pf2id and pf2iw
instructions, which do return the minimum value that fits in
an int32 if the input float is a large negative number.
Fixes: https://bugs.launchpad.net/qemu/+bug/1815423
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-Id: <20190805180332.10185-1-peter.maydell@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The SSE4.1 phminposuw instruction finds the minimum 16-bit element in
the source vector, putting the value of that element in the low 16
bits of the destination vector, the index of that element in the next
three bits and zeroing the rest of the destination. The helper for
this operation fills the destination from high to low, meaning that
when the source and destination are the same register, the minimum
source element can be overwritten before it is copied to the
destination. This patch fixes it to fill the destination from low to
high instead, so the minimum source element is always copied first.
This fixes one gcc test failure in my GCC 6-based testing (and so
concludes the present sequence of patches, as I don't have any further
gcc test failures left in that testing that I attribute to QEMU bugs).
Signed-off-by: Joseph Myers <joseph@codesourcery.com>
Message-Id: <alpine.DEB.2.20.1708111422580.11919@digraph.polyomino.org.uk>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
One of the cases of the SSE4.2 pcmpestri / pcmpestrm / pcmpistri /
pcmpistrm instructions does a substring search. The implementation of
this case in the pcmpxstrx helper is incorrect. The operation in this
case is a search for a string (argument d to the helper) in another
string (argument s to the helper); if a copy of d at a particular
position would run off the end of s, the resulting output bit should
be 0 whether or not the strings match in the region where they
overlap, but the QEMU implementation was wrongly comparing only up to
the point where s ends and counting it as a match if an initial
segment of d matched a terminal segment of s. Here, "run off the end
of s" means that some byte of d would overlap some byte outside of s;
thus, if d has zero length, it is considered to match everywhere,
including after the end of s. This patch fixes the implementation to
correspond with the proper instruction semantics. This fixes four gcc
test failures in my GCC 6-based testing.
Signed-off-by: Joseph Myers <joseph@codesourcery.com>
Message-Id: <alpine.DEB.2.20.1708102139310.8101@digraph.polyomino.org.uk>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The SSE4.1 packusdw instruction combines source and destination
vectors of signed 32-bit integers into a single vector of unsigned
16-bit integers, with unsigned saturation. When the source and
destination are the same register, this means each 32-bit element of
that register is used twice as an input, to produce two of the 16-bit
output elements, and so if the operation is carried out
element-by-element in-place, no matter what the order in which it is
applied to the elements, the first element's operation will overwrite
some future input. The helper for packssdw avoids this issue by
computing the result in a local temporary and copying it to the
destination at the end; this patch fixes the packusdw helper to do
likewise. This fixes three gcc test failures in my GCC 6-based
testing.
Signed-off-by: Joseph Myers <joseph@codesourcery.com>
Message-Id: <alpine.DEB.2.20.1708100023050.9262@digraph.polyomino.org.uk>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
