This CL adds generic SSA rules to propagate constants through raw bits
conversions between floats and integers. This allows constants to
propagate through some math functions. For example, math.Copysign(0, -1)
is now constant folded to a load of -0.0.
Requires a fix to the ARM assembler which loaded -0.0 as +0.0.
Change-Id: I52649a4691077c7414f19d17bb599a6743c23ac2
Reviewed-on: https://go-review.googlesource.com/62250
Run-TryBot: Michael Munday <mike.munday@ibm.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Cherry Zhang <cherryyz@google.com>
Change "yshl" and "yshb" immediate oclass from Yi32 to Yu8.
This forbids:
- negative shift counts
- shift counts that not fit into 8bit
Affects:
RCL{B,L,Q,W}
RCR{B,L,Q,W}
ROL{B,L,Q,W}
ROR{B,L,Q,W}
SAL{B,L,Q,W}
SAR{B,L,Q,W}
SHL{B,L,Q,W}
SHR{B,L,Q,W}
Issue #21528 has some additional context about this change.
Change-Id: I60884cb2b41a860820889fcd878ca6f564006b4a
Reviewed-on: https://go-review.googlesource.com/62190
Run-TryBot: Iskander Sharipov <iskander.sharipov@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
The isStar and directory function parameters have been unused ever since
they were introduced. Remove them.
While at it, apply some other minor simplifications, such as simplifying
a HasPrefix if and using an early continue to unindent many lines of
code.
Change-Id: I8d57353e9ec10cdb59c5388cf6152ce0ec17bdba
Reviewed-on: https://go-review.googlesource.com/62030
Run-TryBot: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Rob Pike <r@golang.org>
Reviewed-by: Marvin Stenger <marvin.stenger94@gmail.com>
There shouldn't be any problems setting error's "Orig" (underlying)
type to a separate anonymous interface, as this is already how
go/types defines it.
Change-Id: I44e9c4048ffe362ce329e8306632e38b5ccfecff
Reviewed-on: https://go-review.googlesource.com/61790
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
Reviewed-by: Alan Donovan <adonovan@google.com>
This adds the VFMADD[213|231]SD, VFNMADD[213|231]SD,
VADDSD, VSUBSD instructions
This will allow us to write a fast path for exp_amd64.s where
these optimizations can be applied in a lot of places.
Change-Id: Ide292107ab887bd1e225a1ad60880235b5ed7c61
Reviewed-on: https://go-review.googlesource.com/61810
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Now that all functions have a DW_AT_frame_base defined we can use
DW_OP_fbreg to specify the location of variables and formal parameters,
instead of the DW_OP_call_frame_cfa/DW_OP_consts/DW_OP_plus, saving 2
bytes for every variable and 2 bytes for every formal parameter after
the first one.
Change-Id: I2c7395b67e4a814a0131ab1520df11ca48ff9327
Reviewed-on: https://go-review.googlesource.com/60550
Run-TryBot: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Heschi Kreinick <heschi@google.com>
DWARF version 4 allows DW_AT_high_pc to be represented as a constant
offset from DW_AT_low_pc, this can help save up to 7 bytes per
function/lexical scope.
Change-Id: I93638d83638ecad4d0d1bfe27348eae6139820c9
Reviewed-on: https://go-review.googlesource.com/60530
Run-TryBot: Heschi Kreinick <heschi@google.com>
Reviewed-by: Heschi Kreinick <heschi@google.com>
3rd change out of 3 to cover AMD64 SSSE3 instruction set in Go asm.
This commit adds instruction that do require new ytab variable.
Change-Id: I0bc7d9401c9176eb3760c3d59494ef082e97af84
Reviewed-on: https://go-review.googlesource.com/56870
Run-TryBot: Iskander Sharipov <iskander.sharipov@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
Reviewed-by: Russ Cox <rsc@golang.org>
This is the last instruction I found missing in SSE2 set.
It does not reuse 'yprefetch' ytabs due to differences in
operands SRC/DST roles:
- PREFETCHx: ModRM:r/m(r) -> FROM
- CLFLUSH: ModRM:r/m(w) -> TO
unaryDst map is extended accordingly.
Change-Id: I89e34ebb81cc0ee5f9ebbb1301bad417f7ee437f
Reviewed-on: https://go-review.googlesource.com/56833
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
Reviewed-by: Russ Cox <rsc@golang.org>
instructions
1st change out of 3 to cover AMD64 SSSE3 instruction set in Go asm.
This commit adds instructions that do not require new named ytab sets.
Change-Id: I0c3dfd8d39c3daa8b7683ab163c63145626d042e
Reviewed-on: https://go-review.googlesource.com/56834
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ilya Tocar <ilya.tocar@intel.com>
Reviewed-by: Russ Cox <rsc@golang.org>
runtime.makemap will allocate map buckets on the heap for hints larger
than the number of elements a single map bucket can hold.
Do not allocate any map bucket on the stack if it is known at compile time
that hint is larger than the number of elements one map bucket can hold.
This avoids zeroing and reserving memory on the stack that will not be used.
Change-Id: I1a5ab853fb16f6a18d67674a77701bf0cf29b550
Reviewed-on: https://go-review.googlesource.com/60450
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Previously the type was first set to uint8 and then corrected
later in walkrange.
Change-Id: I9e4b597710e8a5fad39dde035df85676bc8d2874
Reviewed-on: https://go-review.googlesource.com/61032
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Remove the runtime ismapkey check from makemap and
add a check that the map key type supports comparison
to the hmap construction in the compiler.
Move the ismapkey check for the reflect code path
into reflect_makemap.
Change-Id: I718f79b0670c05b63ef31721e72408f59ec4ae86
Reviewed-on: https://go-review.googlesource.com/61035
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Currently, we handle "x op= y" by rewriting as "x = x op y", while
ensuring that any calls or receive operations in 'x' are only
evaluated once. Notably, pointer indirection, indexing operations,
etc. are left alone as it's typically safe to re-evaluate those.
However, those operations were interleaved with evaluating 'y', which
could include function calls that might cause re-evaluation to yield
different memory addresses.
As a fix, simply ensure that we order side-effecting operations in 'y'
before either evaluation of 'x'.
Fixes#21687.
Change-Id: Ib14e77760fda9c828e394e8e362dc9e5319a84b2
Reviewed-on: https://go-review.googlesource.com/60091
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Do the similar thing to CL 55143 to reduce IMUL.
Change-Id: I1bd38f618058e3cd74fac181f003610ea13f2294
Reviewed-on: https://go-review.googlesource.com/56252
Run-TryBot: Emmanuel Odeke <emm.odeke@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
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>
When compiling a plugin, package main gets a new name so as not to
conflict with the main package in the host binary, or any other
plugins. It is already defined by cmd/go, and used by cmd/link when
filling out the "" package placeholder in symbols.
With this CL, the plugin-specific name for main is also passed to
cmd/compile's -p flag. This is used to fill out the pkgpath field
of types, and ensures that two types defined in two different plugin
mains with the same name will not be mistaken for one another at
runtime.
Fixes#21386
Change-Id: I8a646d8d7451caff533fe0007343ea8b8e1704ed
Reviewed-on: https://go-review.googlesource.com/60910
Run-TryBot: David Crawshaw <crawshaw@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
The "build flags" mentioned in the documentation are only those
that apply to analyzing packages and executing the tool.
Fixes#21711.
Change-Id: Ie7b2a354f1e30c928b40888c51fc68e599a5444a
Reviewed-on: https://go-review.googlesource.com/60830
Reviewed-by: Ian Lance Taylor <iant@golang.org>
The approach of https://golang.org/cl/43476 turned out incorrect.
The problem is that the sniff introduced by the CL only work for simple
expression. And when it fails it fallback to uint64, not int64, which
breaks backward compatibility.
In this CL, we use DWARF for guessing kind instead. That should be more
reliable than previous approach. And importanly, it fallbacks to int64 even
if it fails to guess kind.
Fixes#21708
Change-Id: I39a18cb2efbe4faa9becdcf53d5ac68dba180d46
Reviewed-on: https://go-review.googlesource.com/60510
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
On OpenBSD -current, clang is available/installed as 'cc'. This means that
the existing clang check fails and the clang related flags are not enabled.
Fix this by enabling the clang flags if the compiler claims to support them.
Change-Id: I84d124d2409a10f87002c6cbfdb69b4c9a55981a
Reviewed-on: https://go-review.googlesource.com/60750
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Simplify the DWARF representation of structs by emitting field offsets
as constants rather than location descriptions.
This was not explicitly mentioned as an option in DWARF2. It is
mentioned in DWARF4, but isn't listed in the changes, so it's not clear
if this was always intended to work or is an undocumented change. Either
way, it should be valid DWARF4.
Change-Id: Idf7fdd397a21c8f8745673ecc77ef65afa3ffe1c
Reviewed-on: https://go-review.googlesource.com/51611
Run-TryBot: Heschi Kreinick <heschi@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Since these are type aliases in Go 1.9 make rewriting
golang.org/x/net/context imports to context the default.
Change-Id: Ib7ee58b59ea5463b92e1590c8b8f5c0ef87d310b
Reviewed-on: https://go-review.googlesource.com/58590
Reviewed-by: Sameer Ajmani <sameer@golang.org>
Run-TryBot: Sam Whited <sam@samwhited.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Normal shift rules plus constant folding are enough to generate
efficient shift-by-constant instructions.
Add test to make sure we don't generate comparisons for constant
shifts.
TODO: there are still constant shift rules on PPC64. If they
are removed, the constant folding rules are not enough to remove
all the test and mask stuff for constant shifts. Leave them in
for now.
Fixes#20663.
Change-Id: I724cc324aa8607762d0c8aacf9bfa641bda5c2a1
Reviewed-on: https://go-review.googlesource.com/60330
Run-TryBot: Cherry Zhang <cherryyz@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Currently, cgo supports only macros which can be reduced to constants
or variables. The CL addresses remaining parts, macros which can be
represented as niladic functions.
The basic idea is simple:
1. make a thin wrapper function per macros.
2. replace macro expansions with function calls.
Fixes#10715Fixes#18720
Change-Id: I150b4fb48e9dc4cc34466ef6417c04ac93d4bc1a
Reviewed-on: https://go-review.googlesource.com/43970
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Some instructions operating on <= 32 bits also zero out upper 32bits.
Remove zeroextensions of such values. Triggers a few times during
all.bash. Also removes ugly code like:
MOVL CX,CX
Change-Id: I66a46c190dd6929b7e3c52f3fe6b967768d00638
Reviewed-on: https://go-review.googlesource.com/58090
Run-TryBot: Ilya Tocar <ilya.tocar@intel.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
This makes the name of the function to construct the map bucket type
consistent with runtimes naming and the existing hmap function.
Change-Id: If4d8b4a54c92ab914d4adcb96022b48d8b5db631
Reviewed-on: https://go-review.googlesource.com/59915
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
Padding needed for map buckets is dependent on the types used to
construct the map bucket. In case of indirect keys or values pointers
are used in the map bucket to the keys or values.
Change the map bucket padding calculation to take the alignment of
the key and value types used to construct the map bucket into account
instead of the original key and value type.
Since pointers are always 32bit aligned on amd64p32 this prevents
adding unneeded padding in case the key or value would have needed
64bit alignment without indirect referencing.
Change-Id: I7943448e882d269b5cff7e921a2a6f3430c50878
Reviewed-on: https://go-review.googlesource.com/60030
Reviewed-by: Keith Randall <khr@golang.org>
Check map invariants, type size and alignments during compile time.
Keep runtime checks for reflect by adding them to reflect_makemap.
Change-Id: Ia28610626591bf7fafb7d5a1ca318da272e54879
Reviewed-on: https://go-review.googlesource.com/59914
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
* rename to emitPcln because I'd like to skip not only container types,
but also something like "go.buildid" in the future.
* return bool instead of int.
Change-Id: I029adb81292f7dd2fe98e69f3877c5c27f32ec30
Reviewed-on: https://go-review.googlesource.com/59415
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Change-Id: I8a54235c8b7bf1010f19d0d358cd4f76fc911d28
Reviewed-on: https://go-review.googlesource.com/59416
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Change-Id: Ifc66bd8c32e788f35da3bd105bcdcfd04bd3dabc
Reviewed-on: https://go-review.googlesource.com/59414
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
* extract pkgname() and findlib() from the function for #18190.
* rename const pkgname to const pkgdef to avoid confliction.
Change-Id: Ie62509bfbddcf19cf92b5b12b598679a069e6e74
Reviewed-on: https://go-review.googlesource.com/59417
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Right now we only kind of sort of trace GC STW events. We emit events
around mark termination, but those start well after stopping the world
and end before starting it again, and we don't emit any events for
sweep termination.
Fix this by generalizing EvGCScanStart/EvGCScanDone. These were
already re-purposed to indicate mark termination (despite the names).
This commit renames them to EvGCSTWStart/EvGCSTWDone, adds an argument
to indicate the STW reason, and shuffles the runtime to generate them
right before stopping the world and right after starting the world,
respectively.
These events will make it possible to generate precise minimum mutator
utilization (MMU) graphs and could be useful in detecting
non-preemptible goroutines (e.g., #20792).
Change-Id: If95783f370781d8ef66addd94886028103a7c26f
Reviewed-on: https://go-review.googlesource.com/55411
Reviewed-by: Rick Hudson <rlh@golang.org>
