Robert Griesemer [Thu, 2 Sep 2021 23:43:29 +0000 (16:43 -0700)]
cmd/compile/internal/types2: detect constraint type inference cycles
See the detailed explanations in the code.
Fixes #48136.
Change-Id: I1667aabfbbff97967913b080c77e7ec04ea82feb
Reviewed-on: https://go-review.googlesource.com/c/go/+/347300
Trust: Robert Griesemer <gri@golang.org>
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
Joel Sing [Mon, 9 Aug 2021 08:14:04 +0000 (08:14 +0000)]
cmd/internal/obj/riscv: improve code generation for loading of constants
Loading of constants that are 12 bits or smaller is currently performed using a single
ADDIW instruction, while constants between 13 bits and 32 bits are loaded using a
LUI+ADDIW pair.
Instead, use a single ADDI instruction for the 12 bits or smaller case - this
translates to the LI pseudo-instruction, making objdump more readable and giving:
In the case where a constant exceeds 12 bits, an LUI instruction is required,
however if the lower 12 bits are zero, the ADDIW instruction can be omitted.
The same applies to the case where immediate splitting is performed for other
immediate instructions.
This removes around 900 instructions from the Go binary.
Change-Id: Id6c77774b3b429fa525da018a6926b85df838a2f
Reviewed-on: https://go-review.googlesource.com/c/go/+/344457
Trust: Joel Sing <joel@sing.id.au>
Run-TryBot: Joel Sing <joel@sing.id.au>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com>
Joel Sing [Fri, 20 Aug 2021 17:04:35 +0000 (17:04 +0000)]
cmd/internal/obj/riscv: avoid obj.Prog rewriting for address to register loads
Rather than rewriting the obj.Prog for a MOV pseudo-instruction targeting
an address to register load, generate the appropriate machine instruction
sequence directly.
Change-Id: I507195b6d24ff3253eadcc807ddbe27dca97d220
Reviewed-on: https://go-review.googlesource.com/c/go/+/344456
Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com>
Joel Sing [Thu, 19 Aug 2021 08:35:12 +0000 (08:35 +0000)]
cmd/internal/obj/riscv: avoid obj.Prog rewriting for immediate splitting
Rather than rewriting the obj.Prog for a immediate instructions that need
splitting, generate the appropriate machine instruction sequence directly.
Change-Id: Ie90f0e2a98f97a29281e445c4c3b0c47b793ef4d
Reviewed-on: https://go-review.googlesource.com/c/go/+/344453
Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com>
Joel Sing [Tue, 17 Aug 2021 18:25:05 +0000 (18:25 +0000)]
cmd/internal/obj/riscv: avoid obj.Prog rewriting for memory stores
Rather than rewriting the obj.Prog for a MOV pseudo-instruction targeting
a register to memory stores, generate the appropriate machine instruction
sequence directly.
Change-Id: I6eac8637e1fcb48c04d9f331cac7ae5d8cc54092
Reviewed-on: https://go-review.googlesource.com/c/go/+/344455
Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com>
Joel Sing [Fri, 19 Mar 2021 14:09:59 +0000 (14:09 +0000)]
cmd/internal/obj/riscv: avoid obj.Prog rewriting for memory loads
Rather than rewriting the obj.Prog for a MOV pseudo-instruction targeting
a memory to register load, generate the appropriate machine instruction
sequence directly.
Change-Id: I4c7292ba00f576ec71d4842b6ff27a8ce6db0650
Reviewed-on: https://go-review.googlesource.com/c/go/+/344454
Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com>
Joel Sing [Thu, 19 Aug 2021 05:33:01 +0000 (05:33 +0000)]
cmd/internal/obj/riscv: avoid obj.Prog rewriting for store instructions
Rather than rewriting the obj.Prog for a store instruction, generate the
appropriate machine instruction sequence directly.
Change-Id: I026250ca3d249a1013243948a4e567b708b72d0c
Reviewed-on: https://go-review.googlesource.com/c/go/+/344452
Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Joel Sing [Wed, 18 Aug 2021 18:14:52 +0000 (18:14 +0000)]
cmd/internal/obj/riscv: avoid obj.Prog rewriting for load instructions
Rather than rewriting the obj.Prog for a load instruction, generate the
appropriate machine instruction sequence directly.
Change-Id: Iba656dbf2dff2bce02aa221d9a5cad7b7c4630d8
Reviewed-on: https://go-review.googlesource.com/c/go/+/344451
Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Joel Sing [Tue, 17 Aug 2021 18:45:11 +0000 (18:45 +0000)]
cmd/internal/obj/riscv: avoid obj.Prog rewriting for constant loads
Rather than rewriting the obj.Prog for a MOV pseudo-instruction targeting
a constant to register load, generate the appropriate machine instruction
sequence directly.
Change-Id: I38e62f282b39be2a0a241f32280d306558d49b44
Reviewed-on: https://go-review.googlesource.com/c/go/+/344450
Trust: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Current wasm_exec.js does not enable ECMAScript strict mode. But it is
recommended to be enabled because it
1. eliminates some ECMAScript silent errors by changing them to throw
errors
2. fixes mistakes that make it difficult for JavaScript engines to
perform optimizations
3. prohibits some syntax likely to be defined in future versions of
ECMAScript
This commit enables ECMAScript strict mode in wasm_exec.js following the
transition guide:
Change-Id: Ib8ffceee37e9127698fb51304241f1e429efe83e
Reviewed-on: https://go-review.googlesource.com/c/go/+/334269 Reviewed-by: Richard Musiol <neelance@gmail.com> Reviewed-by: Cherry Mui <cherryyz@google.com>
Trust: Richard Musiol <neelance@gmail.com>
Run-TryBot: Cherry Mui <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Joel Sing [Mon, 23 Aug 2021 15:01:06 +0000 (01:01 +1000)]
cmd/internal/obj/riscv: factor out instructions for MOV pseudo-instructions
This factors out the machine instruction generation for MOV pseudo-instructions,
which will simplify further changes.
Change-Id: Ic0d2c3ae9e0881f7894af50ed45e93b0e4961632
Reviewed-on: https://go-review.googlesource.com/c/go/+/344461
Trust: Joel Sing <joel@sing.id.au>
Trust: Meng Zhuo <mzh@golangcn.org>
Run-TryBot: Joel Sing <joel@sing.id.au>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Meng Zhuo <mzh@golangcn.org> Reviewed-by: Cherry Mui <cherryyz@google.com>
Joe Sylve [Thu, 2 Sep 2021 19:09:15 +0000 (19:09 +0000)]
debug/dwarf: check for DWARFv4 AttrDataBitOffset value
AttrBitOffset is deprecated (but reserved) in DWARFv4. This fix adds
logic to check the new AttrDataBitOffset attribute if AttrBitOffset
attribute is not present.
Fixes #46784
Change-Id: I7406dcaa4c98e95df72361fd4462c39e6be8879d
GitHub-Last-Rev: 5aa10d04910a09538320b4de8fbd8a1f5fd8c17d
GitHub-Pull-Request: golang/go#46790
Reviewed-on: https://go-review.googlesource.com/c/go/+/328709
Run-TryBot: Than McIntosh <thanm@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Trust: Michael Knyszek <mknyszek@google.com>
Matthew Dempsky [Thu, 2 Sep 2021 19:14:47 +0000 (12:14 -0700)]
test: only test -G=3 by default
-G=0 is in maintenance mode, so limit testing it to the longtest
builders.
Change-Id: Ie8a01866b506183d0201f2a3730377cfa663da80
Reviewed-on: https://go-review.googlesource.com/c/go/+/347298
Trust: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Thu, 2 Sep 2021 14:50:42 +0000 (10:50 -0400)]
go/types: systematic detection of missing instantiation
This is a port of CL 346471 to go/types. Additionally, CheckExpr was
updated for the new API to explicitly allow generic expressions.
The error messages in issue39634.go2 are different because go/parser
produces an IndexExpr with BadExpr index value, for backward
compatibility.
Change-Id: I725926de183a016381513fe0e750d1280688ce29
Reviewed-on: https://go-review.googlesource.com/c/go/+/347391
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Damien Neil [Fri, 6 Aug 2021 02:26:21 +0000 (19:26 -0700)]
net/http: fix hang in probing for a zero-length request body
Fix a hang that occurs when making a request and all of the following apply:
* The request method is one of GET, HEAD, DELETE, OPTIONS, PROPFIND, or SEARCH.
* The Request.Body is non-nil.
* The content length is not set, or is set to -1.
* Transfer-Encoding: chunked is not set.
* The request body does not respond to a read within 200ms.
In this case, we give up on probing for a zero-length body and send the
request while the probe completes in the background. Fix a bug in the
io.Reader wrapping the in-flight probe: It should return io.EOF after
the probe completes, but does not.
Fixes #47568.
Change-Id: I7f9188c96e1210055df68424081af927006e4816
Reviewed-on: https://go-review.googlesource.com/c/go/+/340256
Trust: Damien Neil <dneil@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Ingo Oeser <nightlyone@googlemail.com> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Damien Neil [Tue, 22 Jun 2021 03:23:36 +0000 (20:23 -0700)]
net/http: close request body after recovering from a handler panic
When recovering from a panic in a HTTP handler, close the request body
before closing the *conn, ensuring that the *conn's bufio.Reader is safe
to recycle.
Fixes #46866.
Change-Id: I3fe304592e3b423a0970727d68bc1229c3752939
Reviewed-on: https://go-review.googlesource.com/c/go/+/329922
Trust: Damien Neil <dneil@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Dmitri Shuralyov [Thu, 2 Sep 2021 00:04:30 +0000 (20:04 -0400)]
all: update vendored dependencies for Go 1.18
Go 1.18 development is well underway. This is a time to update all
golang.org/x/... module versions that contribute packages to the std
and cmd modules in the standard library to latest master versions.
gotip $ updatestd -goroot=$(pwd) -branch=master
> go version
go version devel go1.18-2872496ba5 Wed Sep 1 23:41:53 2021 +0000 darwin/amd64
> go env GOROOT
/Users/dmitshur/gotip
> go version -m /Users/dmitshur/go/bin/bundle
/Users/dmitshur/go/bin/bundle: go1.17
path golang.org/x/tools/cmd/bundle
mod golang.org/x/tools v0.1.5 h1:ouewzE6p+/VEB31YYnTbEJdi8pFqKp4P4n85vwo3DHA=
dep golang.org/x/mod v0.4.2 h1:Gz96sIWK3OalVv/I/qNygP42zyoKp3xptRVCWRFEBvo=
dep golang.org/x/sys v0.0.0-20210510120138-977fb7262007 h1:gG67DSER+11cZvqIMb8S8bt0vZtiN6xWYARwirrOSfE=
dep golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 h1:go1bK/D/BFZV2I8cIQd1NKEZ+0owSTG1fDTci4IqFcE=
updating bundles in /Users/dmitshur/gotip/src
> go generate -run=bundle std cmd
The x/tools module will be updated in a following CL,
after issue #48124 is resolved.
The module in GOROOT/src/crypto/ed25519/internal/edwards25519/field/_asm
directory is not updated in this CL.
For #36905.
Change-Id: I728000e8465c0fbf6976629e6da42cc4f9be20fc
Reviewed-on: https://go-review.googlesource.com/c/go/+/347191
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Carlos Amedee <carlos@golang.org>
Trust: Dmitri Shuralyov <dmitshur@golang.org>
Robert Findley [Thu, 2 Sep 2021 14:36:37 +0000 (10:36 -0400)]
go/internal/gcimporter: always call SetTParams, even if empty
This resolves an outstanding TODO to be consistent about calling
SetTParams.
Done in both go/internal/gcimporter and cmd/compile/internal/importer.
Change-Id: Ief642efe016f02bab1bab960f6731a0ac0a8ad28
Reviewed-on: https://go-review.googlesource.com/c/go/+/347389
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Also add the new golang.org/x/crypto/curve25519/internal/field package
(it was created in x/crypto CL 315269) to TestDependencies in go/build.
Position it next to its upstream copy, since its required dependencies
are expected to be identical.
For #36905.
Change-Id: I589499cd7176c9b4b8758e59025653d19b58130e
Reviewed-on: https://go-review.googlesource.com/c/go/+/347190
Trust: Dmitri Shuralyov <dmitshur@golang.org>
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Carlos Amedee <carlos@golang.org>
Keith Randall [Wed, 1 Sep 2021 15:33:42 +0000 (08:33 -0700)]
cmd/link: mark stacks as non-executable on freebsd
Fixes #48112
Change-Id: I673a5fe61171e5ba3d8ce58daac79391fb4758f9
Reviewed-on: https://go-review.googlesource.com/c/go/+/346872
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Change-Id: I53d6ba03246372b6f7abd5ac57dbb71715d6eb4a
Reviewed-on: https://go-review.googlesource.com/c/go/+/347189
Run-TryBot: Dmitri Shuralyov <dmitshur@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Carlos Amedee <carlos@golang.org>
Trust: Dmitri Shuralyov <dmitshur@golang.org>
Than McIntosh [Thu, 26 Aug 2021 19:59:36 +0000 (15:59 -0400)]
cmd/link: avoid crash on undefined func sym with external linking
Fix a buglet in relocation processing that crops up with external
linking when you have an undefined function symbol that also has a
prototype (as if it were being defined in assembly src).
Fixes #47993.
Change-Id: Ib655492a63b205ffdc124cfd0cb7f7b731571821
Reviewed-on: https://go-review.googlesource.com/c/go/+/345473
Trust: Than McIntosh <thanm@google.com>
Run-TryBot: Than McIntosh <thanm@google.com> Reviewed-by: Austin Clements <austin@google.com> Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Jeremy Faller <jeremy@golang.org>
Ben Shi [Thu, 2 Sep 2021 00:06:42 +0000 (00:06 +0000)]
cmd/internal/obj/riscv: simplify addition with constant
This CL simplifies riscv addition (add r, imm) to
(ADDI (ADDI r, imm/2), imm-imm/2) if imm is in specific ranges.
(-4096 <= imm <= -2049 or 2048 <= imm <= 4094)
There is little impact to the go1 benchmark, while the total
size of pkg/linux_riscv64 decreased by about 11KB.
Change-Id: I236eb8af3b83bb35ce9c0b318fc1d235e8ab9a4e
GitHub-Last-Rev: a2f56a07635344a40d6b8a9571f236743122be34
GitHub-Pull-Request: golang/go#48110
Reviewed-on: https://go-review.googlesource.com/c/go/+/346689
Run-TryBot: Ben Shi <powerman1st@163.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Joel Sing <joel@sing.id.au> Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Keith Randall <khr@golang.org>
Trust: Michael Munday <mike.munday@lowrisc.org>
Robert Findley [Wed, 1 Sep 2021 20:51:17 +0000 (16:51 -0400)]
go/internal/gcimporter: add import tests for type parameters
Add a new test TestImportTypeparamTests that compiles and imports
packages contained in test/typeparam, and compares the resulting package
scope with the scope produced by type-checking directly.
In the process, fix a bug in go/types affecting embedded instances with
more than one type argument. This was uncovered by smoketest.go.
To enable this new test it was easiest to move gcimporter_test.go to an
external test, which required copying the pkgExts variable.
Fixes #48101
Change-Id: Ie4d981bf463e886a8d141809805d184dbbf64607
Reviewed-on: https://go-review.googlesource.com/c/go/+/347070
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Robert Findley [Wed, 1 Sep 2021 18:15:05 +0000 (14:15 -0400)]
go/internal/gcimporter: add support for importing parameterized types
Port the necessary logic to go/internal/gcimporter from
cmd/compile/internal/importer/iimport.go to support type parameters.
This is a partial port of several compiler CLs: at least CL 319930,
CL 322609, CL 323029, CL 338192, CL 340251, and CL 340989. Because these
ports were not interleaved with the corresponding go/types API changes,
it is easier to just take the latest importer logic.
Notably, the equivalent of types2.AsTypeParam is not used. It should be
unnecessary.
Updates #48101
Change-Id: I938bd8debc3f6a68a3ad8d44c61ef9c5038be7e5
Reviewed-on: https://go-review.googlesource.com/c/go/+/347069
Trust: Robert Findley <rfindley@google.com>
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Dan Scales <danscales@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
Hans [Wed, 28 Jul 2021 05:57:25 +0000 (05:57 +0000)]
runtime: ensure the fixalloc object size is valid
Usually, fixalloc is used to allocate small, persistent and reuseable
objects. The size is typically between range [sizeof(mlink), _FixAllocChunk].
It's rare for being out of the range. But if it did happen, we got a
hard-to-discover memory corruption. This commit prevents that situation by limiting object's size.
Leonard Wang [Sat, 28 Aug 2021 16:53:00 +0000 (00:53 +0800)]
cmd/compile: workaround inlining of closures with range statements
ORANGE is still not inlineable now. This CL is correct only when the range statement is statically dead, and thus not counted during the inline budget check.
If we support range statements in inlining closures in the future, may require additional processing.
Fixes #48033.
Change-Id: I28f5755c28cfa27e41daef9eff2ae332059909bb
Reviewed-on: https://go-review.googlesource.com/c/go/+/345436
Trust: Cherry Mui <cherryyz@google.com> Reviewed-by: Keith Randall <khr@golang.org>
syscall: drop fallback to accept in Accept on Linux
The minimum required Linux kernel version for Go 1.18 will be changed to
2.6.32, see #45964. The current minimum required version is 2.6.23 and
accept4 was added in 2.6.28, so the fallback to accept in Accept on
Linux can be removed.
syscall: drop fallback to utimes in UtimesNano on Linux
The minimum required Linux kernel version for Go 1.18 will be changed to
2.6.32, see #45964. The current minimum required version is 2.6.23 and
utimensat was added in 2.6.22, so the fallback isn't even necessary for
the current minimum supported version. Remove the fallback to utimes.
os, syscall: remove fallback to pipe syscall on Linux
The minimum required Linux kernel version for Go 1.18 will be changed to
2.6.32, see #45964. The pipe2 syscall was added in 2.6.27, so the
fallback to use the pipe syscall in os.Pipe and syscall.forkExecPipe on
Linux can be removed.
Cuong Manh Le [Wed, 9 Jun 2021 07:26:36 +0000 (14:26 +0700)]
src: emit warning when make.bash run with invalid GOROOT_BOOTSTRAP
When setting an invalid GOROOT_BOOTSTRAP, make.bash try looking at env to
find a valid go command and use that GOROOT instead. It's better if the
script emit a warning, instead of silently change the GOROOT_BOOTSTRAP.
Change-Id: I6038093d5e8c072da27bcae44b125f6774f809a7
Reviewed-on: https://go-review.googlesource.com/c/go/+/326030
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
test: always run test in unified IR mode regardless of explicit -G flag
Currently, if a test explicitly specify -G flag, if that flag conflict
with compiler default -G flag, the test will be skipped. That's the
reason CL 346469 haven't fixed the unified IR stuff, but still make the
unified builder passed.
This CL makes run.go always run the test in unified IR mode, regardless
of the explicit -G flag specified.
Updates #48094
Change-Id: I959ecaff6aca07844f0ffcf36caa60cf0747e8ce
Reviewed-on: https://go-review.googlesource.com/c/go/+/347271
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
CL 346469 added support for unsafe.Sizeof and friends to operate on
generic parameters for compiler importer/exporter. This CL adds support
for unified IR.
Updates #48094
Change-Id: I63af1a7c3478f59b03ecc23229ac2254d3457868
Reviewed-on: https://go-review.googlesource.com/c/go/+/346769
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Robert Griesemer [Thu, 2 Sep 2021 00:01:26 +0000 (17:01 -0700)]
cmd/compile: disable type list syntax for the compiler
Add (temporary) syntax.AllowTypeLists mode to control the
acceptance of type lists; the compiler doesn't set it,
but existing syntax and types2 tests do so that the code
remains exercised while it exists.
Adjust various tests to use the type set notation.
Change-Id: I798e607912552db6bfe38a7cd4324b74c6bf4d95
Reviewed-on: https://go-review.googlesource.com/c/go/+/347249
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
Robert Griesemer [Wed, 1 Sep 2021 00:43:18 +0000 (17:43 -0700)]
cmd/compile/internal/types2: systematic detection of missing instantiation
When type-checking expressions, detection of uninstantiated generic
functions and types was somewhat ad-hoc.
Add an extra parameter "allowGenerics" to rawExpr. If not set, the
result operand cannot be generic.
The only place where rawExpr is called with allowGenerics != false
is from exprOrType, which passes an allowGenerics parameter through.
The only place where exprOrType is called with allowGenerics == true
is when handling index expressions and calls. Make sure that we only
accept generic operands where expected, and check the other branches.
As a result, a recently added varType call (CL 345970) can be removed
again.
This also fixes a bug where an error for a conversion to generic
type was reported after the conversion (i.e., with the converted
value, rather than the generic type). Added a test case for that.
For #48048.
Change-Id: I8576326f5fcfb58d78b3ce8572068aa32e66c568
Reviewed-on: https://go-review.googlesource.com/c/go/+/346471
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
Damien Neil [Wed, 1 Sep 2021 19:07:08 +0000 (12:07 -0700)]
all: update golang.org/x/net to pull in CL 346890
For #42777.
Change-Id: I963db8c666e8bcf0fc4f390b359db6408a0f792b
Reviewed-on: https://go-review.googlesource.com/c/go/+/347010
Trust: Damien Neil <dneil@google.com>
Run-TryBot: Damien Neil <dneil@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com>
Matthew Dempsky [Wed, 1 Sep 2021 19:42:35 +0000 (12:42 -0700)]
cmd/compile: optimize unified IR handling of imported functions
This CL skips reading of imported function bodies except from the
local package or for inlining. Previously, we wasted a bunch of CPU
and memory by reading in normal function bodies for all imported
functions and methods.
Change-Id: I0b3df3f4a3e2819117dac11a1f9e4526288c14d8
Reviewed-on: https://go-review.googlesource.com/c/go/+/347030
Trust: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Leonard Wang [Tue, 6 Jul 2021 15:39:29 +0000 (23:39 +0800)]
cmd/compile: remove useless fcount
Change-Id: Ibc23b43dc9a7fabb27d7991977b283459f3deae3
Reviewed-on: https://go-review.googlesource.com/c/go/+/333009
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Michael Munday [Wed, 17 Feb 2021 15:00:34 +0000 (15:00 +0000)]
cmd/{asm,compile}: add fused multiply-add support on riscv64
Add support to the assembler for F[N]M{ADD,SUB}[SD] instructions.
Argument order is:
OP RS1, RS2, RS3, RD
Also, add support for the FMA intrinsic to the compiler. Automatic
FMA matching is left to a future CL.
Change-Id: I47166c7393b2ab6bfc2e42aa8c1a8997c3a071b3
Reviewed-on: https://go-review.googlesource.com/c/go/+/293030
Trust: Michael Munday <mike.munday@lowrisc.org>
Run-TryBot: Michael Munday <mike.munday@lowrisc.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Joel Sing <joel@sing.id.au>
cmd/compile: fix irgen mis-handling invalid function declaration
In -G=3 mode, irgen use its own generated IR, which is mis-handling of
bodyless function and declared function with //go:noescape pragma.
Fix this by adopting the same logic in noder.funcDecl, which minor
change in linkname detection.
Fixes #48097
Change-Id: Ibef921c1f75e071ca61685e0cb4543f2ee5efc7f
Reviewed-on: https://go-review.googlesource.com/c/go/+/346470
Trust: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Cuong Manh Le <cuong.manhle.vn@gmail.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cherry Mui <cherryyz@google.com> Reviewed-by: Dan Scales <danscales@google.com> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
Matthew Dempsky [Tue, 31 Aug 2021 20:21:11 +0000 (13:21 -0700)]
cmd/compile/internal/pkginit: separate "init" and "inittask" logic
This CL splits the creation of the "init" function responsible for
executing package-scope variable initialization statemens from the
creation of the "inittask" record that tells the runtime how to
sequence all program-wide package initialization.
Longer term, this is desirable because sorting variable initialization
is already handled by types2 (with Info.InitOrder), so we might as
well reuse that.
As a more immediate impetus, for unified IR, I want to defer method
wrapper generation until after inlining (to know which wrappers are
needed). But the staticinit optimization used to decide whether to
emit the inittask calls into reflectdata, which in turn tries to
generate its own method wrappers. So separating the work allows to
create the "init" function early and then emit "inittask" after
inlining is done.
Change-Id: Ice1d421f92feecaaeafdf7da6b9647c0f27e3571
Reviewed-on: https://go-review.googlesource.com/c/go/+/346629
Trust: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Paul E. Murphy [Tue, 9 Mar 2021 22:55:18 +0000 (16:55 -0600)]
cmd/internal/obj/ppc64: improve long conditional branch fixup
Improve the code which fixes up conditional branches which exceed the
range of a single instruction by inserting one extra jump when
possible instead of two.
Change-Id: Ib0eb5b0f47f7d0e0ccd55471307a5f73fbda88a9
Reviewed-on: https://go-review.googlesource.com/c/go/+/342930
Run-TryBot: Paul Murphy <murp@ibm.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Lynn Boger <laboger@linux.vnet.ibm.com> Reviewed-by: Cherry Mui <cherryyz@google.com>
korzhao [Mon, 16 Aug 2021 14:25:47 +0000 (22:25 +0800)]
cmd/compile: fix conversions from TypeParam to interface
If the TypeParam has all the methods of an interface, allow conversions from TypeParam to interface
Fixes #47708
Change-Id: I40a82a31f6ea9354130dbe3bcfc83537094bf12c
Reviewed-on: https://go-review.googlesource.com/c/go/+/342509 Reviewed-by: Dan Scales <danscales@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
Trust: Dan Scales <danscales@google.com>
Trust: Keith Randall <khr@golang.org>
Trust: Robert Griesemer <gri@golang.org>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Joel Sing [Thu, 19 Aug 2021 04:06:55 +0000 (04:06 +0000)]
cmd/asm: adjust riscv64 test data to avoid churn
Rather than jumping and branching to a label (the offset for which changes
when instructions are added or removed), use PC-relative offsets. This
reduces unnecessary churn in the instruction encodings.
Change-Id: I6816ce939eeabdf828039f59f4f29260eb1ac8da
Reviewed-on: https://go-review.googlesource.com/c/go/+/344449
Trust: Joel Sing <joel@sing.id.au>
Run-TryBot: Joel Sing <joel@sing.id.au>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Michael Munday <mike.munday@lowrisc.org>
Joel Sing [Sat, 21 Aug 2021 08:40:20 +0000 (18:40 +1000)]
cmd/asm,cmd/internal/obj/riscv: add more error tests for riscv64 assembly
Add more error tests for riscv64 assembly. Also avoid a panic when one of
these error conditions is hit.
Change-Id: If5d913894facbd67f7d014eab745da77c8c66ab0
Reviewed-on: https://go-review.googlesource.com/c/go/+/344228
Trust: Joel Sing <joel@sing.id.au>
Run-TryBot: Joel Sing <joel@sing.id.au>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Meng Zhuo <mzh@golangcn.org>
Keith Randall [Tue, 31 Aug 2021 21:53:53 +0000 (14:53 -0700)]
cmd/compile: assign results of transformAssign back to source location
Otherwise the modifications of transformAssign are dropped on the floor.
Change-Id: Id40782564952ed53f9ade1dba4e85290c8522abc
Reviewed-on: https://go-review.googlesource.com/c/go/+/346590
Trust: Keith Randall <khr@golang.org>
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Dan Scales <danscales@google.com>
Robert Findley [Tue, 31 Aug 2021 22:32:21 +0000 (18:32 -0400)]
go/types: disallow aliases for generic types
This is a port of CL 346294 to go/types.
Change-Id: Ib70541a92e352c8df8123c8b82bb4eeedce3b89f
Reviewed-on: https://go-review.googlesource.com/c/go/+/346560
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 22:26:16 +0000 (18:26 -0400)]
go/types: remove superfluous ordinaryType calls
This is a port of CL 346291 to go/types.
Change-Id: I8f864aca5cdb4037bc27a81cde1597430b9a48db
Reviewed-on: https://go-review.googlesource.com/c/go/+/346559
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 22:20:49 +0000 (18:20 -0400)]
go/types: more systematic error handling in typeWriter
This is a port of CL 346009 to go/types. An unnecessary break statement
was removed from both the port and types2.
Change-Id: I2cc1328a61100d4b01a2d26ac7bac9044440d579
Reviewed-on: https://go-review.googlesource.com/c/go/+/346558
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 22:15:53 +0000 (18:15 -0400)]
go/types: eliminate typeHashing global variable
This is a port of CL 345929 to go/types. It is also a step toward making
instantiation concurrency-safe.
Also fix some whitespace in instantiate.go.
Updates #47910
Change-Id: Icdeb227cb83eee15da6db90daab294c8c55db601
Reviewed-on: https://go-review.googlesource.com/c/go/+/346557
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 22:12:15 +0000 (18:12 -0400)]
go/types: generalize instanceHash to accept any type, rename to typeHash
This is a port of CL 345791 to go/types.
Change-Id: I673c22ad8b668f07aae4117555b1c0efb273fb78
Reviewed-on: https://go-review.googlesource.com/c/go/+/346556
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 21:59:05 +0000 (17:59 -0400)]
go/types: don't print instance markers for type hashes
This is a port of CL 345891 to go/types.
Change-Id: I5abcb9c9c5110923a743f0c47d9b34b2baabab68
Reviewed-on: https://go-review.googlesource.com/c/go/+/346555
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 21:49:53 +0000 (17:49 -0400)]
go/types: use a typeWriter to write types (cleanup)
This is a port of CL 345890 to go/types.
Change-Id: I98162deaf044b2194b05dc51e6948e227216fc4d
Reviewed-on: https://go-review.googlesource.com/c/go/+/346554
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 21:09:47 +0000 (17:09 -0400)]
go/types: do not declare new methods on instantiated types
This is a port of CL 345472 to go/types.
Change-Id: I0e2a88909ecebe9dea3325244153f5c74e4c3ce7
Reviewed-on: https://go-review.googlesource.com/c/go/+/346553
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Change-Id: Icad5fb8b3b4375182f420a51c80607b88696561e
Reviewed-on: https://go-review.googlesource.com/c/go/+/346552
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 21:00:59 +0000 (17:00 -0400)]
go/types: remove need for the instance struct
This is a port of CL 345177 to go/types.
Change-Id: I79fcfbf5b28e9a7a2e66c81bc831e164a8da8bbb
Reviewed-on: https://go-review.googlesource.com/c/go/+/346551
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Robert Findley [Tue, 31 Aug 2021 20:56:07 +0000 (16:56 -0400)]
go/types: address some TODOs (cleanup)
This is a port of CL 345176 to go/types, though not all TODOs were
present in go/types.
A TODO that still needs to be resolved was added back to types2.
Change-Id: Icf79483c92d0bc1248de772c7044620f0f0a5c58
Reviewed-on: https://go-review.googlesource.com/c/go/+/346550
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Keith Randall [Tue, 31 Aug 2021 21:09:41 +0000 (14:09 -0700)]
cmd/compile: use the zero value for results of impossible indexing
type A [0]int
var a A
x := a[i]
Use the zero value for x instead of the "impossible" value. That lets
us at least compile code like this with -B, even though it can't
possibly run correctly.
Fixes #48092
Change-Id: Idad5cfab49e05f375c069b05addceed68a15299f
Reviewed-on: https://go-review.googlesource.com/c/go/+/346589
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: David Chase <drchase@google.com>
TryBot-Result: Go Bot <gobot@golang.org>
Keith Randall [Wed, 25 Aug 2021 20:03:30 +0000 (13:03 -0700)]
cmd/compile: use right line number for conversion expression
Use the position of the actual conversion operation instead of base.Pos.
Fixes #47880
Change-Id: I56adc134e09cb7fd625adc0a847c1a6b3e254b1a
Reviewed-on: https://go-review.googlesource.com/c/go/+/345095
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com>
These helper functions are no longer necessary, now that type parameters
are enabled; we can access type parameters directly.
When considering the existence or non-existence of type parameters, we
can either check whether node.TParams != nil, or whether
node.TParams.NumFields() > 0. The heuristic I'm using for deciding
between these checks is as follows:
- For data access, just check node.TParams != nil.
- For producing errors if type parameters exist, check NumFields() > 0.
Change-Id: I6597536898e975564e9e8bf6a3a91bc798e0f110
Reviewed-on: https://go-review.googlesource.com/c/go/+/346549
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 18:57:53 +0000 (14:57 -0400)]
go/types: fix type set printing and add a test
This is a port of CL 344873 to go/types.
Change-Id: Iad41d6de166214178adf2123ac05023895f5208b
Reviewed-on: https://go-review.googlesource.com/c/go/+/346435
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Findley [Tue, 31 Aug 2021 18:57:24 +0000 (14:57 -0400)]
go/types: rename IsMethodSet to IsConstraint (cleanup)
This is a port of CL 344872 to go/types.
Change-Id: Id794f1fc3b86779ee32dbe5e656ffc747d44c3e2
Reviewed-on: https://go-review.googlesource.com/c/go/+/346434
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Robert Findley [Tue, 31 Aug 2021 18:41:29 +0000 (14:41 -0400)]
go/types: add error reporting for 1.18 syntax if GoVersion is below 1.18
This is a port of CL 344871 to go/types. Unlike the compiler, go/parser
is already always producing 1.18 syntax, so the effect of this CL is to
add some additional errors when Config.GoVersion is below 1.18.
This is a non-trivial port, both due to different error reporting APIs
and due to interacting with declaration syntax nodes, which differ
between go/ast and cmd/compile/internal/syntax.
Change-Id: I8003a014e6eec5e554c24e9a6cfc0548ec534834
Reviewed-on: https://go-review.googlesource.com/c/go/+/346433
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Robert Findley [Tue, 31 Aug 2021 18:03:33 +0000 (14:03 -0400)]
go/types: allow composite literals of type parameter type
This is a port of CL 342690 to go/types.
Change-Id: I27dcde237e400a84c3394a3579805014777830bc
Reviewed-on: https://go-review.googlesource.com/c/go/+/346432
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Dan Scales [Fri, 13 Aug 2021 01:39:24 +0000 (18:39 -0700)]
cmd/compile: fix handling of Defn field during stenciling
When the Defn field of a name node is not an ONAME (for a closure
variable), then it points to a body node of the same function/closure.
Therefore, we should not attempt to substitute it at the time we are
substituting the local variables. Instead, we remember a mapping from the
Defn node to the nodes that reference it, and update the Defn fields of
the copied name nodes at the time that we create the new copy of the
Defn node.
Added some comments to the Defn field of ir.Name.
Moved the Defn (and Outer code, for consistency) from namelist() to
localvar(), since Defn needs to updated for all local variables, not
just those in a closure. Fixed case where .Defn was not being set
properly in noder2 for type switches. Fixed another case where the Defn
field had to be updated during transformSelect() because the Defn node
was being completely changed to a new node.
Fixed some spacing in typeswitch2.go
Fixes #47676
Fixes #48016
Change-Id: Iae70dd76575f4a647c1db79e1eba9bbe44bfc226
Reviewed-on: https://go-review.googlesource.com/c/go/+/346290
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Dan Scales <danscales@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
Martin Möhrmann [Tue, 31 Aug 2021 17:41:33 +0000 (19:41 +0200)]
cmd/compile: remove folding of 32 bit pointer offsets on amd64
These rules were likely only used in the removed amd64p32 port.
Passes toolstash -cmp.
Change-Id: Ie626d1021bade7f7571d1bd5271fbe8676bb295e
Reviewed-on: https://go-review.googlesource.com/c/go/+/346509
Trust: Martin Möhrmann <martin@golang.org>
Run-TryBot: Martin Möhrmann <martin@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Martin Möhrmann [Mon, 30 Aug 2021 20:42:17 +0000 (22:42 +0200)]
cmd/compile: add MOVOstoreconst with offset folding on amd64
Replace MOVOstorezero with new MOVOstoreconst.
MOVOstoreconst has similar address folding rules then
other MOVstoreconst operations but only supports zero
as store value. Currently only MOVO stores with zero
values are generated. Using MOVOstoreconst with
SymValAndOff aux has the advantage that we can just
add one more MOVstoreconst variant to the existing rules.
The main effect of this CL is converting 16 byte zeroing
of a value on the stack from LEAQ+MOVUPS to just MOVUPS
which reduces binary size.
Change-Id: I08f5dd5ab9251448a4572d3ddd1e0c8cd417f5e3
Reviewed-on: https://go-review.googlesource.com/c/go/+/346249
Run-TryBot: Martin Möhrmann <moehrmann@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
Trust: Martin Möhrmann <martin@golang.org>
Keith Randall [Tue, 31 Aug 2021 16:31:56 +0000 (09:31 -0700)]
cmd/compile: builtins may be in the unsafe package
Now that unsafe.Sizeof and friends can operate on generic parameters,
and evaluate to non-constants, we need to export/import them correctly.
Fixes #48094
Change-Id: If3ebf77255385cd5462e13fb7ced8b157ba3cf5b
Reviewed-on: https://go-review.googlesource.com/c/go/+/346469
Trust: Keith Randall <khr@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org> Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Robert Griesemer [Tue, 31 Aug 2021 03:30:25 +0000 (20:30 -0700)]
cmd/compile/internal/types2: disallow aliases for generic types
The existing approach (alias name stands for generic type name)
is an exception: it's the only place where a generic type could
be used without explicit instantiation. The correct solution is
currently under discussion (see proposal issue #46477).
This CL requires that the RHS of an alias type declaration be
an instantiated non-generic type. If #46477 is accepted, the
implementation will require proper representation of alias
types.
Change-Id: Ie85b923213a64f39837e56e38e14757458272b93
Reviewed-on: https://go-review.googlesource.com/c/go/+/346294
Trust: Robert Griesemer <gri@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
The value types in type assertions and type switches cannot be
constraint types (if there are, an error was reported earlier),
so there is no need to check again that they are not constraint
types.
This permits merging the ordinaryType call with varType, which
is the only place where it's needed.
Change-Id: I44a852377b3dddf53692f764e588801fb3d3c0a8
Reviewed-on: https://go-review.googlesource.com/c/go/+/346291
Trust: Robert Griesemer <gri@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
Matthew Dempsky [Mon, 30 Aug 2021 08:01:20 +0000 (01:01 -0700)]
cmd/compile: make unified IR more selective about method wrappers
This CL makes two changes to how unified IR emits method wrappers:
1. It no longer emits wrappers for defined types' underlying
types. Previously, a declaration like `type T struct { U }` would emit
wrappers for both `T` and `struct { U }`. Now they're only emitted for
`T`.
2. It emits method value wrappers only when OMETHVALUE nodes are
actually created, like how -G=0 works. Method values are relatively
rare, aren't needed for runtime type descriptors (unlike method
expression wrappers), and large projects end up spending a non-trivial
amount of time compiling these unneeded wrappers.
Change-Id: I21da97df3132ec12cc67debf62b5b2d282f481cf
Reviewed-on: https://go-review.googlesource.com/c/go/+/346230
Trust: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Cuong Manh Le <cuong.manhle.vn@gmail.com>
Robert Findley [Mon, 30 Aug 2021 21:40:45 +0000 (17:40 -0400)]
go/types, types2: add a test for invalid import of "init"
This error reporting code path did not have test coverage, and panics in
1.17 (filed as #48082).
Add a test that would have reproduced the panic, for both go/types and
cmd/compile/internal/types2.
Change-Id: Icd5f54f8407e4ab57d432f44a129ecf6b2755feb
Reviewed-on: https://go-review.googlesource.com/c/go/+/346309
Trust: Robert Findley <rfindley@google.com>
Run-TryBot: Robert Findley <rfindley@google.com> Reviewed-by: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org>
Robert Griesemer [Sun, 29 Aug 2021 17:57:06 +0000 (10:57 -0700)]
cmd/compile/internal/types2: more systematic error handling in typeWriter
When using a typeWriter for debugging/error message type strings,
it shouldn't crash in the presence of type-checker internal bugs.
But when a typeHasher is used, we don't want to silently ignore
errors.
Introduce an error method that panics in type hashing mode but
prints an error value otherwise.
Also fixed an incorrect 'if' statement in tParamList.
Change-Id: I26c8b8e0b14396e91ad71bf903e36ce1ca55839e
Reviewed-on: https://go-review.googlesource.com/c/go/+/346009
Trust: Robert Griesemer <gri@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
Robert Griesemer [Sat, 28 Aug 2021 03:50:31 +0000 (20:50 -0700)]
cmd/compile/internal/types2: generalize instanceHash to accept any type, rename to typeHash
Rename instanceHashing accordingly.
Eventually, this will make it possible to use typeHash to detect
multiple identical types in type switch cases and other places.
Also fix some bugs: When creating a type hash, the name of function
parameters must be ignored because they don't matter for type
identity. And when printing a type name, don't assume its type
is a *Named type; it could be a *Basic type as well.
Finally, use a correctly qualified type string when reporting
a duplicate type error in a type switch case rather than the
(debugging) type string.
Change-Id: Ida3873f6259b51847843b0e2d7e3aa2fcdc3a0c3
Reviewed-on: https://go-review.googlesource.com/c/go/+/345791
Trust: Robert Griesemer <gri@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
Quim Muntal [Sun, 29 Aug 2021 20:57:36 +0000 (22:57 +0200)]
go/types, types2: types in method expressions must be instantiated
Use varType instead of instantiatedOperand to check if the type of a method expressions is instantiated.
This removes the last usage of instantiatedOperand, so it can be deleted.
Fixes #48048
Change-Id: I2b219dafe2bba3603100bb8f25b8ff4e8ef53841
Reviewed-on: https://go-review.googlesource.com/c/go/+/345970
Trust: Robert Griesemer <gri@golang.org>
Trust: Matthew Dempsky <mdempsky@google.com>
Run-TryBot: Robert Griesemer <gri@golang.org>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Dan Scales [Sun, 29 Aug 2021 13:49:28 +0000 (06:49 -0700)]
cmd/compile: fix bug with Inferred targs
We were using the type from the wrong Node (the partially filled-in
FUNCINST) rather than the original function node - which is pointed to
by the OFUNCINST)) to set the final fully-substituted type of the
OFUNCINST. So fixed the node reference. Also, added check so we don't do
any work at all if the OFUNCINST already has all type args filled in.
Added few extra cases to the test file issue48030.go, to cover
fully-specified type args, partially inferred type args, and fully
inferred type args.
Fixes #48030
Change-Id: If9e4f2e0514d68b9d241f30c423259133932b25b
Reviewed-on: https://go-review.googlesource.com/c/go/+/346229
Trust: Dan Scales <danscales@google.com>
Run-TryBot: Dan Scales <danscales@google.com>
TryBot-Result: Go Bot <gobot@golang.org> Reviewed-by: Keith Randall <khr@golang.org>
Robert Griesemer [Sun, 29 Aug 2021 00:11:51 +0000 (17:11 -0700)]
cmd/compile/internal/types2: don't print instance markers for type hashes
Since we know whether we are printing a type string used as
instance hash, don't print instance markers, so that we don't
need to remove them afterwards either.
Change-Id: Ib01627b6da989ef89d51e734810a3377eb466925
Reviewed-on: https://go-review.googlesource.com/c/go/+/345891
Trust: Robert Griesemer <gri@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>
Robert Griesemer [Sat, 28 Aug 2021 20:00:41 +0000 (13:00 -0700)]
cmd/compile/internal/types2: use a typeWriter to write types (cleanup)
Rather then passing through a buffer, qualified, and visited
list to each helper function, maintain state in a typeWriter
object and use methods on it. This cleans up this code quite
a bit.
Use a map ("seen") for cycle detection rather than a list.
Move printing of [ and ] for type lists into the typeList
method so that callers don't have to do it themselves.
Change-Id: I1346373e979cb90710fbc073953aa51e2f6581f9
Reviewed-on: https://go-review.googlesource.com/c/go/+/345890
Trust: Robert Griesemer <gri@golang.org> Reviewed-by: Robert Findley <rfindley@google.com>