HMUL is commutative. However, it has asymmetric register requirements.
There are existing rewrite rules to place arguments in preferable slots.
Due to a bug, the existing rulegen commutativity engine doesn't generate
the commuted form of the HMUL rules.
The commuted form of those rewrite rules cause infinite loops.
In order to fix the rulegen commutativity bug,
we need to choose between eliminating
those rewrite rules and marking HMUL ops as not commutative.
This change chooses the latter, since doing so yields better
optimization results on std+cmd.
Removing the rewrite rules yields only text size regressions:
file before after Δ %
runtime.s 477257 477269 +12 +0.003%
time.s 83552 83612 +60 +0.072%
encoding/asn1.s 57378 57382 +4 +0.007%
cmd/go/internal/modfetch/codehost.s 89822 89829 +7 +0.008%
cmd/internal/test2json.s 9459 9466 +7 +0.074%
cmd/go/internal/test.s 57665 57678 +13 +0.023%
Marking HMUL as not commutative actually yields (mostly) improvements:
file before after Δ %
runtime.s 477257 477247 -10 -0.002%
math.s 35985 35992 +7 +0.019%
strconv.s 53486 53462 -24 -0.045%
syscall.s 82483 82446 -37 -0.045%
time.s 83552 83561 +9 +0.011%
os.s 52691 52684 -7 -0.013%
archive/zip.s 42285 42272 -13 -0.031%
encoding/asn1.s 57378 57329 -49 -0.085%
encoding/base64.s 12156 12094 -62 -0.510%
net.s 296286 296276 -10 -0.003%
encoding/base32.s 9720 9658 -62 -0.638%
net/http.s 560931 560907 -24 -0.004%
net/smtp.s 14421 14411 -10 -0.069%
cmd/vendor/golang.org/x/sys/unix.s 74307 74266 -41 -0.055%
The regressions are minor, and are in functions math.cbrt,
time.Time.String, and time.Date.
Change-Id: I9f6d9ee71654e5b70381cac77b0ac26011f4ea12
Reviewed-on: https://go-review.googlesource.com/c/go/+/213701
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
This adds asynchronous preemption function for amd64 and 386. These
functions spill and restore all register state that can be used by
user Go code.
For the moment we stub out the other arches.
For #10958, #24543.
Change-Id: I6f93fabe9875f4834922a5712362e79045c00aca
Reviewed-on: https://go-review.googlesource.com/c/go/+/201759
Run-TryBot: Austin Clements <austin@google.com>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
For #10958, #24543, but makes sense on its own.
Change-Id: I2a87dab66b82a1863e4b6512b1f8def51463ce2a
Reviewed-on: https://go-review.googlesource.com/c/go/+/203284
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
To permit ssa-level optimization, this change introduces an amd64 intrinsic
that generates the VFMADD231SD instruction for the fused-multiply-add
operation on systems that support it. System support is detected via
cpu.X86.HasFMA. A rewrite rule can then translate the generic ssa intrinsic
("Fma") to VFMADD231SD.
The benchmark compares the software implementation (old) with the intrinsic
(new).
name old time/op new time/op delta
Fma-4 27.2ns ± 1% 1.0ns ± 9% -96.48% (p=0.008 n=5+5)
Updates #25819.
Change-Id: I966655e5f96817a5d06dff5942418a3915b09584
Reviewed-on: https://go-review.googlesource.com/c/go/+/137156
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
A few examples (for accessing a slice of length 3):
s[-1] runtime error: index out of range [-1]
s[3] runtime error: index out of range [3] with length 3
s[-1:0] runtime error: slice bounds out of range [-1:]
s[3:0] runtime error: slice bounds out of range [3:0]
s[3:-1] runtime error: slice bounds out of range [:-1]
s[3:4] runtime error: slice bounds out of range [:4] with capacity 3
s[0:3:4] runtime error: slice bounds out of range [::4] with capacity 3
Note that in cases where there are multiple things wrong with the
indexes (e.g. s[3:-1]), we report one of those errors kind of
arbitrarily, currently the rightmost one.
An exhaustive set of examples is in issue30116[u].out in the CL.
The message text has the same prefix as the old message text. That
leads to slightly awkward phrasing but hopefully minimizes the chance
that code depending on the error text will break.
Increases the size of the go binary by 0.5% (amd64). The panic functions
take arguments in registers in order to keep the size of the compiled code
as small as possible.
Fixes#30116
Change-Id: Idb99a827b7888822ca34c240eca87b7e44a04fdd
Reviewed-on: https://go-review.googlesource.com/c/go/+/161477
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Refactoring only.
This makes it easier to add ops
that do indexed memory loads/stores.
Passes toolstash-check.
Change-Id: I82df0d4154718577ec42106fa1bc76571bf65096
Reviewed-on: https://go-review.googlesource.com/c/go/+/166425
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
The current support_XXX variables are specific for the
amd64 and 386 platforms.
Prefix processor capability variables by architecture to have a
consistent naming scheme and avoid reuse of the existing
variables for new platforms.
This also aligns naming of runtime variables closer with internal/cpu
processor capability variable names.
Change-Id: I3eabb29a03874678851376185d3a62e73c1aff1d
Reviewed-on: https://go-review.googlesource.com/c/91435
Run-TryBot: Martin Möhrmann <martisch@uos.de>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
prove is able to find 94 occurrences in std cmd where a divisor
can't have the value -1. The change removes
the extraneous fix-up code for these cases.
Fixes#25239
Change-Id: Ic184de971f47cc57c702eb72805b8e291c14035d
Reviewed-on: https://go-review.googlesource.com/c/130215
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Add generic, 386 and amd64 specific ops and SSA rules for multiplication
with overflow and branching based on overflow flags. Use these to intrinsify
runtime/internal/math.MulUinptr.
On amd64
mul, overflow := math.MulUintptr(a, b)
if overflow {
is lowered to two instructions:
MULQ SI
JO 0x10ee35c
No codegen tests as codegen can not currently test unexported internal runtime
functions.
amd64:
name old time/op new time/op delta
MulUintptr/small 1.16ns ± 5% 0.88ns ± 6% -24.36% (p=0.000 n=19+20)
MulUintptr/large 10.7ns ± 1% 1.1ns ± 1% -89.28% (p=0.000 n=17+19)
Change-Id: If60739a86f820e5044d677276c21df90d3c7a86a
Reviewed-on: https://go-review.googlesource.com/c/141820
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
According to AMD64.rules, BTS&BTR&BTC use arg1 as the bit index,
while BT uses arg0. This CL fixes the wrong comment message in
AMD64Ops.go, which indicates all bit indexes are in arg0.
Change-Id: Idb78f4d39f7ef5ea78065ad8bc651324597e2a8a
Reviewed-on: https://go-review.googlesource.com/135419
Reviewed-by: Keith Randall <khr@golang.org>
Code generation for OpAMD64CMOV[WLQ]EQF uses AX as a scratch register,
but only CMOVQEQF, correctly lets compiler know. Mark other 2 as
clobbering AX.
Fixes#26097
Change-Id: I2a65bd67bf18a540898b4a0ae6c8766e0b767b19
Reviewed-on: https://go-review.googlesource.com/121336
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
Reviewed-by: Giovanni Bajo <rasky@develer.com>
Turns out it doesn't currently matter, as these ops are always issued
together with a BTSL which does clobber flags. So I can't write a test
that currently fails. But better to be future-proof.
BS{F,R}Q generates flags, so it doesn't need to be marked as clobbering.
Change-Id: I70daea154023fd435fac696bf3a384803c647cd3
Reviewed-on: https://go-review.googlesource.com/117375
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Don't do direct loads from argument slots if the sizes don't match.
This prevents us from loading from a float32 using a uint64 load
during expressions like uint64(math.float32Bits(f)) where f is a float32 arg.
Fixes#25322
Change-Id: I3887d76f78c844ba546243e7721d811c3d4a9700
Reviewed-on: https://go-review.googlesource.com/112637
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Some *mem ops are loads, some are stores, some are modifications.
Replace mem->load for the loads.
Replace mem->store for the stores.
Replace mem->modify for the load-modify-stores.
The only semantic change in this CL is to mark
ADD(Q|L)constmodify (which used to be ADD(Q|L)constmem) as
both a read and a write, instead of just a write. This is arguably
a bug fix, but the bug isn't triggerable at the moment, see CL 112157.
Change-Id: Iccb45aea817b606adb2d712ff99b10ee28e4616a
Reviewed-on: https://go-review.googlesource.com/112159
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Currently, each architecture lowers OpConvert to an arch-specific
OpXXXconvert. This is silly because OpConvert means the same thing on
all architectures and is logically a no-op that exists only to keep
track of conversions to and from unsafe.Pointer. Furthermore, lowering
it makes it harder to recognize in other analyses, particularly
liveness analysis.
This CL eliminates the lowering of OpConvert, leaving it as the
generic op until code generation time.
The main complexity here is that we still need to register-allocate
OpConvert operations. Currently, each arch's lowered OpConvert
specifies all GP registers in its register mask. Ideally, OpConvert
wouldn't affect value homing at all, and we could just copy the home
of OpConvert's source, but this can potentially home an OpConvert in a
LocalSlot, which neither regalloc nor stackalloc expect. Rather than
try to disentangle this assumption from regalloc and stackalloc, we
continue to register-allocate OpConvert, but teach regalloc that
OpConvert can be allocated to any allocatable GP register.
For #24543.
Change-Id: I795a6aee5fd94d4444a7bafac3838a400c9f7bb6
Reviewed-on: https://go-review.googlesource.com/108496
Run-TryBot: Austin Clements <austin@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
This reverts commit 080187f4f7.
It broke build of golang.org/x/exp/shiny/iconvg
See issue 24395 for details
Change-Id: Ifd6134f6214e6cee40bd3c63c32941d5fc96ae8b
Reviewed-on: https://go-review.googlesource.com/100755
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
cmd/asm now supports three-operand form of IMUL,
so instead of using IMUL with resultInArg0, emit IMUL3 instruction.
This results in less redundant MOVs where SSA assigns
different registers to input[0] and dst arguments.
Note: these have exactly the same encoding when reg0=reg1:
IMUL3x $const, reg0, reg1
IMULx $const, reg
Two-operand IMULx is like a crippled IMUL3x, with dst fixed to input[0].
This is why we don't bother to generate IMULx for the case where
dst is the same as input[0].
Change-Id: I4becda475b3dffdd07b6fdf1c75bacc82af654e4
Reviewed-on: https://go-review.googlesource.com/99656
Run-TryBot: Iskander Sharipov <iskander.sharipov@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Giovanni Bajo <rasky@develer.com>
Reviewed-by: Keith Randall <khr@golang.org>
Add a bool to opInfo to indicate if an Op never results in any
instructions. This is a conservative approximation: some operations,
like Copy, may or may not generate code depending on their arguments.
I built the list by reading each arch's ssaGenValue function. Hopefully
I got them all.
Change-Id: I130b251b65f18208294e129bb7ddc3f91d57d31d
Reviewed-on: https://go-review.googlesource.com/97957
Reviewed-by: Keith Randall <khr@golang.org>
Allow the compiler to generate code like CMPQ 16(AX), $7
It's tricky because it's difficult to spill such a comparison during
flagalloc, because the same memory state might not be available at
the restore locations.
Solve this problem by decomposing the compare+load back into its parts
if it needs to be spilled.
The big win is that the write barrier test goes from:
MOVL runtime.writeBarrier(SB), CX
TESTL CX, CX
JNE 60
to
CMPL runtime.writeBarrier(SB), $0
JNE 59
It's one instruction and one byte smaller.
Fixes#19485Fixes#15245
Update #22460
Binaries are about 0.15% smaller.
Change-Id: I4fd8d1111b6b9924d52f9a0901ca1b2e5cce0836
Reviewed-on: https://go-review.googlesource.com/86035
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
The sub-word shifts need to sign-extend before shifting, to avoid
bringing in data from higher in the argument.
Fixes#23812
Change-Id: I0a95a0b49c48f3b40b85765bb4a9bb492be0cd73
Reviewed-on: https://go-review.googlesource.com/93716
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
The caller's PC is always available in the frame. We can just
load it when needed, no need to spill.
Change-Id: I9c0a525903e574bb4eec9fe53cbeb8c64321166a
Reviewed-on: https://go-review.googlesource.com/70710
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Add a compiler intrinsic for getcallersp. So we are able to get
rid of the argument (not done in this CL).
Change-Id: Ic38fda1c694f918328659ab44654198fb116668d
Reviewed-on: https://go-review.googlesource.com/69350
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Reviewed-by: David Chase <drchase@google.com>
Combine setcc and store of result into setcc that writes directly to memory.
Triggers 200+ times in go tool.
Fixes#21630
Change-Id: Iafa22607426f4120140c88fae4b9aecb46e0bba8
Reviewed-on: https://go-review.googlesource.com/67950
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
First step towards removing the mandatory argument for
getcallerpc, which solves certain problems for the runtime.
This might also slightly improve performance.
Intrinsic enabled on 386, amd64, amd64p32,
runtime asm implementation removed on those architectures.
Now-superfluous argument remains in getcallerpc signature
(for a future CL; non-386/amd64 asm funcs ignore it).
Added getcallerpc to the "not a real function" test
in dcl.go, that story is a little odd with respect to
unexported functions but that is not this CL.
Fixes#17327.
Change-Id: I5df1ad91f27ee9ac1f0dd88fa48f1329d6306c3e
Reviewed-on: https://go-review.googlesource.com/31851
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Rematerializable ops can be inserted after the flagalloc phase,
they must therefore not clobber flags. This CL adds a check to
ensure this doesn't happen and fixes the instances where it
does currently.
amd64: ADDQconst and ADDLconst were recently changed to be
rematerializable in CL 54393 (only in tip, not 1.9). That change
has been reverted.
s390x: MOVDaddr could clobber flags when using dynamic linking due
to a ADD with immediate instruction. Change the code generation to
use LA/LAY instead.
Fixes#21080.
Change-Id: Ia85c882afa2a820a309e93775354b3169ec6d034
Reviewed-on: https://go-review.googlesource.com/63030
Run-TryBot: Michael Munday <mike.munday@ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
Currently we only use 1 and 4 as a scale for indexed 4-byte load.
In code generated in #20711 we can use indexed load with scale=8,
to improve performance:
name old time/op new time/op delta
GM-6 108µs ± 0% 95µs ± 0% -12.06% (p=0.000 n=10+10)
So add new ops and combine loadidx1(shift 3..).. into loadidx8,
same for stores.
Change-Id: I5ed1c250ac40960e20606580cf9de221e75b72f1
Reviewed-on: https://go-review.googlesource.com/46134
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
The assembler barfs on large offsets. Make sure that all the
instructions that need to have their offsets in an int32
1) check on any rule that computes offsets for such instructions
2) change their aux fields so the check builder checks it.
The assembler also silently misassembled offsets between 1<<31
and 1<<32. Add a check in the assembler to barf on those as well.
Fixes#21655
Change-Id: Iebf24bf10f9f37b3ea819ceb7d588251c0f46d7d
Reviewed-on: https://go-review.googlesource.com/59630
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
We can rematerialize only ops that have SP or SB as their only argument.
There are some ADDQconst(SP) that can be rematerialized, but are spilled/filled instead,
so mark addconst as rematerializeable. This shaves ~1kb from go tool.
Change-Id: Ib4cf4fe5f2ec9d3d7e5f0f77f1193eba66ca2f08
Reviewed-on: https://go-review.googlesource.com/54393
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Implement int reg <-> fp reg moves on amd64.
If we see a load to int reg followed by an int->fp move, then we can just
load to the fp reg instead. Same for stores.
math.Abs is now:
MOVQ "".x+8(SP), AX
SHLQ $1, AX
SHRQ $1, AX
MOVQ AX, "".~r1+16(SP)
math.Copysign is now:
MOVQ "".x+8(SP), AX
SHLQ $1, AX
SHRQ $1, AX
MOVQ "".y+16(SP), CX
SHRQ $63, CX
SHLQ $63, CX
ORQ CX, AX
MOVQ AX, "".~r2+24(SP)
math.Float64bits is now:
MOVSD "".x+8(SP), X0
MOVSD X0, "".~r1+16(SP)
(it would be nicer to use a non-SSE reg for this, nothing is perfect)
And due to the fix for #21440, the inlined version of these improve as well.
name old time/op new time/op delta
Abs 1.38ns ± 5% 0.89ns ±10% -35.54% (p=0.000 n=10+10)
Copysign 1.56ns ± 7% 1.35ns ± 6% -13.77% (p=0.000 n=9+10)
Fixes#13095
Change-Id: Ibd7f2792412a6668608780b0688a77062e1f1499
Reviewed-on: https://go-review.googlesource.com/58732
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
This prevents unnecessary reg-reg moves during pointer arithmetic.
This change reduces the size of the full hello world binary by 0.4%.
Updates #21572
Change-Id: Ia0427021e5c94545a0dbd83a6801815806e5b12d
Reviewed-on: https://go-review.googlesource.com/58371
Run-TryBot: Josh Bleecher Snyder <josharian@gmail.com>
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
We can add a constant to loaction in memory with 1 instruction,
as opposed to load+add+store, so add a new op and relevent ssa rules.
Triggers in e. g. encoding/json isValidNumber:
NumberIsValid-6 36.4ns ± 0% 35.2ns ± 1% -3.32% (p=0.000 n=6+10)
Shaves ~2.5 kb from go tool.
Change-Id: I7ba576676c2522432360f77b290cecb9574a93c3
Reviewed-on: https://go-review.googlesource.com/54431
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
ADDSDmem comment said f32 (likely a copy-paste mistake).
Also swap ADDSSmem and ADDSDmem positions in the list to uniform the
list order.
Fixes#21225
Change-Id: I26bb116900c1cf4c4e6faeef613d7318c9c85b98
Reviewed-on: https://go-review.googlesource.com/52071
Run-TryBot: Alberto Donizetti <alb.donizetti@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
Reviewed-by: Keith Randall <khr@golang.org>
