Rob Pike [Tue, 8 Sep 2015 21:58:12 +0000 (14:58 -0700)]
text/template: provide a way to trim leading and trailing space between actions
Borrowing a suggestion from the issue listed below, we modify the lexer to
trim spaces at the beginning (end) of a block of text if the action immediately
before (after) is marked with a minus sign. To avoid parsing/lexing ambiguity,
we require an ASCII space between the minus sign and the rest of the action.
Thus:
{{23 -}}
<
{{- 45}}
produces the output
23<45
All the work is done in the lexer. The modification is invisible to the parser
or any outside package (except I guess for noticing some gaps in the input
if one tracks error positions). Thus it slips in without worry in text/template
and html/template both.
Fixes long-requested issue #9969.
Change-Id: I3774be650bfa6370cb993d0899aa669c211de7b2
Reviewed-on: https://go-review.googlesource.com/14391 Reviewed-by: Andrew Gerrand <adg@golang.org>
Michael Hudson-Doyle [Tue, 1 Sep 2015 02:07:14 +0000 (14:07 +1200)]
cmd/dist, cmd/link: force external linking for shared libs on arm
Also run testcshared.
Fixes #12425
Change-Id: I5baea8d772d3462f945aab96260b4197dbb20c0a
Reviewed-on: https://go-review.googlesource.com/14143 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Austin Clements [Sat, 29 Aug 2015 03:15:41 +0000 (23:15 -0400)]
runtime: move stack barrier code to its own file
Currently the stack barrier code is mixed in with the mark and scan
code. Move all of the stack barrier related functions and variables to
a new dedicated source file. There are no code modifications.
Change-Id: I604603045465ef8573b9f88915d28ab6b5910903
Reviewed-on: https://go-review.googlesource.com/14050 Reviewed-by: Rick Hudson <rlh@golang.org>
Michael Hudson-Doyle [Thu, 21 May 2015 01:07:19 +0000 (13:07 +1200)]
cmd/internal/ld: put read-only relocated data into .data.rel.ro when making a shared object
Currently Go produces shared libraries that cannot be shared between processes
because they have relocations against the text segment (not text section). This
fixes this by moving some data to sections with magic names recognized by the
static linker.
The change in genasmsym to add STYPELINK to the switch should fix things on
darwin/arm64.
Fixes #10914
Updates #9210
Change-Id: Iab4a6678dd04cec6114e683caac5cf31b1063309
Reviewed-on: https://go-review.googlesource.com/14306
Run-TryBot: Michael Hudson-Doyle <michael.hudson@canonical.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Just a mechanical copy, no code changes.
This is to reduce code difference when adding the mips64 port.
Change-Id: Id06e975f414a7b09f4827167b30813b228a3bfaf
Reviewed-on: https://go-review.googlesource.com/14324 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Just a mechanical copy with filename renames, no code changes.
This is to reduce code difference when adding the mips64 port.
Change-Id: Id06e975f414a7b09f4827167b30813b228a3bfae
Reviewed-on: https://go-review.googlesource.com/14323 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Michael Hudson-Doyle [Mon, 7 Sep 2015 03:00:52 +0000 (15:00 +1200)]
cmd/internal/obj: remove dead code and small cleanups
Change-Id: I88fa0cc245a2141af04acced8716e08b1133abd1
Reviewed-on: https://go-review.googlesource.com/14350 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Some commits made by Aram from his personal email address are
actually copyright Oracle:
a77fcb3 net: fix comment in sendFile b0e71f4 net: link with networking libraries when net package is in use 92e959a syscall, net: use sendfile on Solaris db8d5b7 net: try to fix setKeepAlivePeriod on Solaris fe5ef5c runtime, syscall: link Solaris binaries directly instead of using dlopen/dlsym 2b90c3e go/build: enable cgo by default on solaris/amd64 2d18ab7 doc/progs: disable cgo tests that use C.Stdout on Solaris 2230e9d misc/cgo: add various solaris build lines 649c7b6 net: add cgo support for Solaris 24396da os/user: small fixes for Solaris 121489c runtime/cgo: add cgo support for solaris/amd64 83b25d9 cmd/ld: make .rela and .rela.plt sections contiguous c94f1f7 runtime: always load address of libcFunc on Solaris e481aac cmd/6l: use .plt instead of .got on Solaris
See bug for clarification.
Fixes #12452
Change-Id: I0aeb1b46c0c7d09c5c736e383ecf40240d2cf85f
Reviewed-on: https://go-review.googlesource.com/14380 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Rob Pike [Tue, 8 Sep 2015 17:58:21 +0000 (10:58 -0700)]
doc: mention that go install removes binaries built by go build
Fixes #12288.
For inclusion in the 1.5.1 release.
Change-Id: I9354b7eaa76000498465c4a5cbab7246de9ecb7c
Reviewed-on: https://go-review.googlesource.com/14382 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Dave Cheney [Sat, 5 Sep 2015 04:43:50 +0000 (14:43 +1000)]
cmd/compile: convert typecheck_stack to []*Node
This one of a set of changes to make the transition away from NodeList
easier by removing cases in which NodeList doesn't act semi-trivially like a
[]*Node.
This CL was originally prepared by Josh Bleecher Snyder <josharian@gmail.com>.
This change passes go build -toolexec 'toolstash -cmp' -a std.
cmd/asm: fix another infinite loop in register list parser
The code parsing register lists involves an inner loop on
each range defined by the lo,hi bounds. The condition on
this loop (for lo<=hi) is fragile, because the bounds
are unsigned 16 bits numbers.
In some corner cases, the calculated upper bound is 2^16-1
leading to an infinite loop.
Parsing operand `[):[o-FP` results in:
- an infinite loop for non ARM architectures
- the generation of almost 2^16 errors for the ARM architecture
(which are then ignored)
This CL improves the code in 3 ways:
- bail out early when parsing non R prefixed registers
- make sure the register index is never negative
- make sure the number of iterations is limited by the
maximum size of the range (as a defensive measure).
Fixes #12469
Change-Id: Ib1e7e36fb8ad5a3a52c50fc6219d3cfe2b39cc34
Reviewed-on: https://go-review.googlesource.com/14314 Reviewed-by: Rob Pike <r@golang.org>
Dave Cheney [Sat, 5 Sep 2015 02:30:13 +0000 (12:30 +1000)]
cmd/compile: use []*Node instead of NodeList in bottomUpVisitor
This one of a set of changes to make the transition away from NodeList
easier by removing cases in which NodeList doesn't act semi-trivially like a
[]*Node.
This CL was originally prepared by Josh Bleecher Snyder <josharian@gmail.com>.
This change passes go build -toolexec 'toolstash -cmp' -a std.
Change-Id: I582ff8b077eb384b84721a1edb0c1efbc0c40059
Reviewed-on: https://go-review.googlesource.com/14304 Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org> Reviewed-by: Dave Cheney <dave@cheney.net>
Run-TryBot: Dave Cheney <dave@cheney.net>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Paul Marks [Sat, 5 Sep 2015 01:33:35 +0000 (18:33 -0700)]
net: Make Listen(":port") use IPv6 when IPv4 is not supported.
When running an experimental kernel with IPv4 disabled, Listen(":port")
currently tries to create an AF_INET socket, and fails. Instead, it
should see !supportsIPv4, and use an AF_INET6 socket.
This sort of environment is quite esoteric at the moment, but I can
force the tests to fail on regular Linux using the following tweaks:
- net/net.go: supportsIPv4, supportsIPv6, supportsIPv4map = false, true, false
- net/sockopt_linux.go: ipv6only=true
- net/ipsock_posix.go: Revert this fix
- ./make.bash && ../bin/go test net
Also, make the arrows in server_test.go point to the left, because
server<-client is easier to read.
Fixes #12510
Change-Id: I0cc3b6b08d5e6908d2fbf8594f652ba19815aa4b
Reviewed-on: https://go-review.googlesource.com/14334
Run-TryBot: Paul Marks <pmarks@google.com> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Dave Cheney [Thu, 3 Sep 2015 07:07:00 +0000 (17:07 +1000)]
cmd/compile: use []*Node instead of NodeList in sinit
This is a first of a set of changes to make the transition away from NodeList
easier by removing cases in which NodeList doesn't act semi-trivially like a
[]*Node.
This CL was originally prepared by Josh Bleecher Snyder <josharian@gmail.com>.
This change passes go build -toolexec 'toolstash -cmp' -a std.
Shawn Walker-Salas [Wed, 26 Aug 2015 22:24:41 +0000 (15:24 -0700)]
cmd/go: fix Go buildid reading on Solaris
TestNoteReading fails on Solaris with linkmode=external due to some
assumptions made about how ELF .note sections are written by some
linkers.
On current versions of Solaris and older derivatives, SHF_ALLOC is
intentionally ignored for .note sections unless the .note section is
assigned to the text segment via a mapfile. Also, if .note sections
are assigned to the text segment, no PT_NOTE program header will be
created thwarting Go's attempts at attempting to quickly find the
.note.
Furthermore, Go assumes that the relevant note segment will be placed
early in the file while the Solaris linker currently places the note
segment last in the file, additionally thwarting Go's optimisation
attempts that read only the first 16KB of the file to find the
buildid.
The fix is to detect when the note section is outside of the first
16KB of the file and then fallback to additionally reading that
section of the file. This way, in future versions of Solaris when
this linking behaviour is changed, the fast path will always succeed
and we'll only be slower if it fails; likewise, any other linker that
does this will also just work.
Change-Id: Ibd2dea475d6ce6a8b4b40e2da19a83fc0514025d
Reviewed-on: https://go-review.googlesource.com/14301 Reviewed-by: Ian Lance Taylor <iant@golang.org>
net/http: optimize some io.Copy calls by reusing buffers
Optimize two calls of io.Copy which cannot make use of neither
io.ReaderFrom nor io.WriterTo optimization tricks by replacing them with
io.CopyBuffer with reusable buffers.
First is fallback call to io.Copy when server misses the optimized case
of using sendfile to copy from a regular file to net.TCPConn; second is
use of io.Copy on piped reader/writer when handler implementation uses
http.CloseNotifier interface. One of the notable users of
http.CloseNotifier is httputil.ReverseProxy.
benchmark old ns/op new ns/op delta
BenchmarkCloseNotifier-4 309591 303388 -2.00%
benchmark old allocs new allocs delta
BenchmarkCloseNotifier-4 50 49 -2.00%
benchmark old bytes new bytes delta
BenchmarkCloseNotifier-4 36168 3140 -91.32%
Michael Hudson-Doyle [Tue, 1 Sep 2015 02:03:38 +0000 (14:03 +1200)]
cmd/compile, cmd/go: generate position independent code with -buildmode=c-shared on arm
All the code was there to do this, it just wasn't hooked up.
Fixes #10914
Change-Id: Ide8f9bbe50fecb5d11cd579915ee98d4c7efe403
Reviewed-on: https://go-review.googlesource.com/14142 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Michael Hudson-Doyle [Fri, 4 Sep 2015 02:57:59 +0000 (14:57 +1200)]
runtime: teach softfloat interpreter about "add r11, pc, r11"
This is generated during fp code when -shared is active.
Change-Id: Ia1092299b9c3b63ff771ca4842158b42c34bd008
Reviewed-on: https://go-review.googlesource.com/14286 Reviewed-by: Ian Lance Taylor <iant@golang.org> Reviewed-by: Dave Cheney <dave@cheney.net>
Shenghou Ma [Tue, 1 Sep 2015 23:58:31 +0000 (19:58 -0400)]
cmd/link/internal/ld: align PE .text section to 32-byte when external linking
Some symbols, for example, masks requires 16-byte alignment, and
they are placed in the text section. Before this change, the text
section is only aligned to 4-byte, and it's making masks unaligned.
Fixes #12415.
Change-Id: I7767778d1b4f7d3e74c2719a02848350782a4160
Reviewed-on: https://go-review.googlesource.com/14166
Run-TryBot: Minux Ma <minux@golang.org> Reviewed-by: Alex Brainman <alex.brainman@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Michael Hudson-Doyle [Thu, 21 May 2015 01:07:19 +0000 (13:07 +1200)]
cmd/internal/ld: put read-only relocated data into .data.rel.ro when making a shared object
Currently Go produces shared libraries that cannot be shared between processes
because they have relocations against the text segment (not text section). This
fixes this by moving some data to sections with magic names recognized by the
static linker.
Fixes #10914
Updates #9210
Change-Id: I7178daadc0ae87953d5a084aa3d580f4e3b46d47
Reviewed-on: https://go-review.googlesource.com/10300
Run-TryBot: Michael Hudson-Doyle <michael.hudson@canonical.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Michael Hudson-Doyle [Thu, 3 Sep 2015 10:49:18 +0000 (22:49 +1200)]
cmd/asm: fix handling of nested #if[n]defs
The lexer needs to process all #if[n]defs, even those found when processing is
disabled by a preceding failed conditional, or the first #endif in something
like:
#ifdef <undefined>
#ifdef whatever
#endif
#endif
terminates the first #ifdef and the second causes an error. And then the
processing of the inner #ifdefs needs to ignore their argument when they are
disabled by an outer failed condition.
Change-Id: Iba259498f1e16042f5b7580b9c000bb0599733d0
Reviewed-on: https://go-review.googlesource.com/14253 Reviewed-by: Rob Pike <r@golang.org>
Joe Tsai [Thu, 3 Sep 2015 22:24:53 +0000 (15:24 -0700)]
hash: update documentation for MakeTable in crc32 and crc64
Explicitly say that *Table returned by MakeTable may not be
modified. Otherwise, this leads to very subtle bugs that may
or may not manifest themselves.
Same comment was made on package crc64, to keep the future
open to the caching tables that crc32 effectively does.
Fixes: #12487.
Change-Id: I2881bebb8b16f6f8564412172774c79c2593c6c1
Reviewed-on: https://go-review.googlesource.com/14258 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Rob Pike [Wed, 2 Sep 2015 23:05:04 +0000 (16:05 -0700)]
net/rpc: don't exit if Accept gets an error
The default implementation of Accept, which spins up a new server
for every new connection, calls log.Fatal if the listener is closed,
stopping any outstanding work. Change that to a non-fatal log
call so work can continue.
There is no programmatic signaling of the problem, just the log,
but that should be enough.
Chris Hines [Tue, 25 Aug 2015 01:48:39 +0000 (21:48 -0400)]
database/sql: close bad connections in commit or rollback:
Previously Tx.close always passed a nil error to tx.db.putConn. As a
result bad connections were reused, even if the driver returned
driver.ErrBadConn. Adding an err parameter to Tx.close allows it to
receive the driver error from Tx.Commit and Tx.Rollback and pass it
to tx.db.putConn.
Michael Hudson-Doyle [Tue, 11 Aug 2015 00:29:00 +0000 (12:29 +1200)]
cmd/internal/obj: some platform independent bits of proper toolchain support for thread local storage
Also simplifies some silliness around making the .tbss section wrt internal
vs external linking. The "make TLS make sense" project has quite a few more
steps to go.
Issue #11270
Change-Id: Ia4fa135cb22d916728ead95bdbc0ebc1ae06f05c
Reviewed-on: https://go-review.googlesource.com/13990 Reviewed-by: Ian Lance Taylor <iant@golang.org> Reviewed-by: David Crawshaw <crawshaw@golang.org>
Run-TryBot: David Crawshaw <crawshaw@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Joe Tsai [Wed, 2 Sep 2015 06:42:43 +0000 (23:42 -0700)]
compress/flate: simplify inflate logic
The flate library contains generator code, which is used to generate
the fixed huffman table. This is done so that fixed blocks can be
processed quicker since there is no need generate the decoder table
for fixed codes.
Instead, delete the precomputed table, and use sync.Once to generate
it at runtime when used.
Advantages:
* Reduces duplicated logic in flate package
* Reduces binary size by approximately 2KiB
Disadvantages:
* For the simplest possible program that simply decodes the fixed
block "\x03\x00" once, the modified code takes 4.7% longer for the
first decode. Compression performance for subsequent blocks afterwards
has no noticeable slow down.
Change-Id: I8f351218debf7d732118808859eda481b01011f6
Reviewed-on: https://go-review.googlesource.com/14181 Reviewed-by: Nigel Tao <nigeltao@golang.org>
Didier Spezia [Wed, 26 Aug 2015 14:27:19 +0000 (14:27 +0000)]
cmd/asm: fix several panics with erroneous input
The parser tries to read as much information as possible,
issuing some errors when needed. Errors generally do not
stop the parsing.
With some pathological input, it may result in various
panics when the error message itself is built, or when the
next operand is parsed. It happens while parsing
pseudo-instructions.
For instance, the following lines all generate a panic:
TEXT
TEXT%
TEXT 1,1
TEXT $"toto", 0, $1
FUNCDATA
DATA 0
DATA(0),1
FUNCDATA(SB
GLOBL 0, 1
PCDATA 1
Added corresponding tests.
Introduced a writer in the parser to capture error messages
for testing purpose. It defaults to os.Stderr.
Added an explicit check when symbol names cannot be displayed.
Interrupted parsing early when the number of operands is wrong for
pseudo-instructions.
Note that the last point is a change of behavior, because some
operands will not get parsed anymore in case of early error.
IMO, it is acceptable, because only the first error of the line
is considered anyway. If it is not acceptable, it can probably
be improved at the price of a more verbose CL.
Fixes #11765
Fixes #11760
Fixes #11759
Change-Id: I9602a848132e358a1bccad794d7555e0823970dd
Reviewed-on: https://go-review.googlesource.com/13925 Reviewed-by: Rob Pike <r@golang.org>
Fabian Wickborn [Thu, 20 Aug 2015 08:45:51 +0000 (10:45 +0200)]
build: Fix bootstrap.bash for official source tarballs
At the moment, bootstrap.bash assumes it is called from a git working
copy. Hence, it fails to complete when running in an unpacked official
source tarball where .git and .gitignore do not exist. This fix adds a
test for existence for .git and a -f switch for the removal of
.gitignore.
Fixes #12223
Change-Id: I7f305b83b38d5115504932bd38dadb7bdeb5d487
Reviewed-on: https://go-review.googlesource.com/13770 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Andrew Gerrand <adg@golang.org>
Nigel Tao [Mon, 31 Aug 2015 03:38:13 +0000 (13:38 +1000)]
image/draw: optimize out some bounds checks.
We could undoubtedly squeeze even more out of these loops, and in the
long term, a better compiler would be smarter with bounds checks, but in
the short term, this small change is an easy win.
benchmark old ns/op new ns/op delta
BenchmarkFillOver-8 16194701323192 -18.29%
BenchmarkCopyOver-8 11293691062787 -5.90%
BenchmarkGlyphOver-8 420070 378608 -9.87%
On github.com/golang/freetype/truetype's BenchmarkDrawString:
benchmark old ns/op new ns/op delta
BenchmarkDrawString-8 95614358807019 -7.89%
Change-Id: Ib1c6271ac18bced85e0fb5ebf250dd57d7747e75
Reviewed-on: https://go-review.googlesource.com/14093 Reviewed-by: Rob Pike <r@golang.org>
Alexander Morozov [Mon, 31 Aug 2015 16:40:25 +0000 (09:40 -0700)]
syscall: remove unused kernelVersion function from tests
Change-Id: If0d00999c58f7421e4da06e1822ba5abccf72cac
Reviewed-on: https://go-review.googlesource.com/14111 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Dave Cheney [Mon, 31 Aug 2015 11:15:29 +0000 (21:15 +1000)]
internal/obj/arm64: remove CASE and BCASE
Fixes #10994
CASE and BCASE were used by 7c in switch statements, cmd/compile
does not use them, cmd/assemble couldn't assemble them, and the arm64
peephole optimiser didn't know about them.
Michael Hudson-Doyle [Mon, 3 Aug 2015 02:08:17 +0000 (14:08 +1200)]
cmd/link: pass value being relocated to archreloc
And clean up the mess on arm64 (the mess on arm is too confusing).
See issue #10050
Change-Id: I2ce813fe8646d4e818eb660612a7e4b2bb04de4c
Reviewed-on: https://go-review.googlesource.com/13884 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Austin Clements [Wed, 26 Aug 2015 19:06:43 +0000 (15:06 -0400)]
runtime: check that stack barrier unwind is in sync
Currently the stack barrier stub blindly unwinds the next stack
barrier from the G's stack barrier array without checking that it's
the right stack barrier. If through some bug the stack barrier array
position gets out of sync with where we actually are on the stack,
this could return to the wrong PC, which would lead to difficult to
debug crashes. To address this, this commit adds a check to the amd64
stack barrier stub that it's unwinding the correct stack barrier.
Austin Clements [Wed, 26 Aug 2015 17:54:26 +0000 (13:54 -0400)]
runtime: add GODEBUG for stack barriers at every frame
Currently enabling the debugging mode where stack barriers are
installed at every frame requires recompiling the runtime. However,
this is potentially useful for field debugging and for runtime tests,
so make this mode a GODEBUG.
Austin Clements [Wed, 26 Aug 2015 16:16:51 +0000 (12:16 -0400)]
runtime: don't install a stack barrier in cgocallback_gofunc's frame
Currently the runtime can install stack barriers in any frame.
However, the frame of cgocallback_gofunc is special: it's the one
function that switches from a regular G stack to the system stack on
return. Hence, the return PC slot in its frame on the G stack is
actually used to save getg().sched.pc (so tracebacks appear to unwind
to the last Go function running on that G), and not as an actual
return PC for cgocallback_gofunc.
Because of this, if we install a stack barrier in cgocallback_gofunc's
return PC slot, when cgocallback_gofunc does return, it will move the
stack barrier stub PC in to getg().sched.pc and switch back to the
system stack. The rest of the runtime doesn't know how to deal with a
stack barrier stub in sched.pc: nothing knows how to match it up with
the G's stack barrier array and, when the runtime removes stack
barriers, it doesn't know to undo the one in sched.pc. Hence, if the C
code later returns back in to Go code, it will attempt to return
through the stack barrier saved in sched.pc, which may no longer have
correct unwinding information.
Fix this by blacklisting cgocallback_gofunc's frame so the runtime
won't install a stack barrier in it's return PC slot.
Adam Langley [Thu, 9 Jul 2015 22:42:02 +0000 (15:42 -0700)]
crypto/tls: reject ServerHellos with empty ALPN protocols.
https://tools.ietf.org/html/rfc7301#section-3.1 specifies that a
ProtocolName may not be empty. This change enforces this for ServerHello
messages—it's already enforced for ClientHello messages.
Vlad Krasnov [Thu, 28 May 2015 20:50:23 +0000 (13:50 -0700)]
crypto/aes: dedicated asm version of AES-GCM
The existing implementation didn't use the CLMUL instructions for fast
and constant time binary-field multiplication. With this change, amd64
CPUs that support both AES and CLMUL instructions will use an optimised
asm implementation.
aubble [Thu, 20 Aug 2015 18:31:15 +0000 (14:31 -0400)]
crypto/tls: allow tls.Listen when only GetCertificate is provided.
Go 1.5 allowed TLS connections where Config.Certificates was nil as long
as the GetCertificate callback was given. However, tls.Listen wasn't
updated accordingly until this change.
Change-Id: I5f67f323f63c988ff79642f3daf8a6b2a153e6b2
Reviewed-on: https://go-review.googlesource.com/13801 Reviewed-by: Adam Langley <agl@golang.org>
Håvard Haugen [Tue, 9 Jun 2015 22:30:32 +0000 (00:30 +0200)]
testing/quick: terminate for arbitrary recursive types
Recursive types R containing slices of R's did not terminate despite the
effort in CL 10821.
For recursive types there was a competition between slice expansion by a
factor 'complexSize', and termination with probability '1/complexSize'
which lead to stack overflow as soon as a recursive struct had slices
pointing to its own type.
Fix this by shrinking the size hint as a function of recursion depth.
This has the dual effect of reducing the number of elements generated
per slice and also increasing the probability for termination.
Fixes #11148.
Change-Id: Ib61155b4f2e2de3873d508d63a1f4be759426d67
Reviewed-on: https://go-review.googlesource.com/13830 Reviewed-by: Adam Langley <agl@golang.org>