Michael Hudson-Doyle [Tue, 5 May 2015 04:10:12 +0000 (16:10 +1200)]
cmd/6l, cmd/internal/ld: handle R_PCREL to function in other shared library
An ELF linker handles a PC-relative reference to an STT_FUNC defined in a
shared library by building a PLT entry and referring to that, so do the
same in 6l.
Fixes #10690
Change-Id: I061a96fd4400d957e301d0ac86760ce256910e1d
Reviewed-on: https://go-review.googlesource.com/9711
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 [Tue, 5 May 2015 02:17:07 +0000 (14:17 +1200)]
cmd/internal/ld: reserve space for package list note when -buildmode=shared
This makes the intermediate object file a little bigger but it doesn't waste
any space in the final shared library.
Fixes #10691
Change-Id: Ic51a571d60291f1ac2dad1b50dba4679643168ae
Reviewed-on: https://go-review.googlesource.com/9710 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Michael Hudson-Doyle [Tue, 14 Apr 2015 10:03:21 +0000 (12:03 +0200)]
cmd/go: rebuild stale shared objects before linking against them.
This changes the action graph when shared libraries are involved to always have
an action for the shared library (which does nothing when the shared library
is up to date).
Change-Id: Ibbc70fd01cbb3f4e8c0ef96e62a151002d446144
Reviewed-on: https://go-review.googlesource.com/8934 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 [Fri, 1 May 2015 00:49:36 +0000 (10:49 +1000)]
cmd/go: fix linux-amd64-clang builder
Fixes #10660
Fix the clang only builder by passing -extld down to the linker when needed.
The build passed on most hosts because gcc is almost always present. The bug
was verified by symlinking bin/false in place of gcc in my $PATH and running
the build.
Also, resolve a TODO and move the support logic into its own function.
Change-Id: I4e27a1119356e295500a0d19ad7a4ec14207bf10
Reviewed-on: https://go-review.googlesource.com/9526
Run-TryBot: Dave Cheney <dave@cheney.net> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Rob Pike [Tue, 5 May 2015 18:17:08 +0000 (11:17 -0700)]
os: add LookupEnv, like Getenv but reports presence
Fixes #9676.
Change-Id: I32fe474cdfa09aff91daa4b10ac4df28ffdaa649
Reviewed-on: https://go-review.googlesource.com/9741 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Rob Pike [Tue, 5 May 2015 18:05:35 +0000 (11:05 -0700)]
fmt: document that Scanf returns an error the same as Scan
No semantic change.
Fixes #8708.
Change-Id: Ieda04a86a19bb69bfc2519d381a2f025e7cb8279
Reviewed-on: https://go-review.googlesource.com/9740 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Rob Pike [Mon, 4 May 2015 20:09:31 +0000 (13:09 -0700)]
text/template: shut down lexing goroutine on error
When a parse error occurred, the lexing goroutine would lay idle.
It's not likely a problem but if the program is for some reason
accepting badly formed data repeatedly, it's wasteful.
The solution is easy: Just drain the input on error. We know this
will succeed because the input is always a string and is therefore
guaranteed finite.
With debugging prints in the package tests I've shown this is effective,
shutting down 79 goroutines that would otherwise linger, out of 123 total.
Keith Randall [Thu, 30 Apr 2015 21:56:35 +0000 (14:56 -0700)]
runtime: let freezetheworld work even when gomaxprocs=1
Freezetheworld still has stuff to do when gomaxprocs=1.
In particular, signals can come in on other Ms (like the GC M, say)
and the single user M is still running.
These a series of changes fix inconsistent errors on the package net
APIs. Now almost all the APIs return OpError as a common error type
except Lookup, Resolve and Parse APIs. The Lookup, Resolve and Parse
APIs return more specific errors such as DNSError, AddrError or
ParseError.
An OpError may contain nested error information. For example, Dial may
return an OpError containing a DNSError, AddrError, unexposed type/value
or other package's type/value like the following:
OpError{/* dial info */, Err: &DNSError{}}
OpError{/* dial info */, Err: &AddrError{}}
OpError{/* dial info */, Err: <unexposed type or value>}
OpError{/* dial info */, Err: <other package's type or value>}
and Read and Write may return an OpError containing other OpError when
an application uses io.Copy or similar:
OpError{/* for io.Reader */, Err: &OpError{/* for io.Writer */}}
When an endpoint is created for connection-oriented byte-stream
protocols, Read may return an io.EOF when the connection is closed by
remote endpoint.
Change-Id: Id678e369088dc9fbe9073cfe7ff8a8754a57d61f
Reviewed-on: https://go-review.googlesource.com/9236 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Mikio Hara [Sat, 2 May 2015 08:58:06 +0000 (17:58 +0900)]
net: add missing ReadFrom, WriteTo deadline tests
Change-Id: If84edfaec361ca2fbb75707c4ad30e4ce64f7013
Reviewed-on: https://go-review.googlesource.com/9664 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Shenghou Ma [Thu, 30 Apr 2015 23:03:31 +0000 (19:03 -0400)]
runtime: fix software FP regs corruption when emulating SQRT on ARM
When emulating ARM FSQRT instruction, the sqrt function itself
should not use any floating point arithmetics, otherwise it will
clobber the user software FP registers.
Fortunately, the sqrt function only uses floating point instructions
to test for corner cases, so it's easy to make that function does
all it job using pure integer arithmetic only. I've verified that
after this change, runtime.stepflt and runtime.sqrt doesn't contain
any call to _sfloat. (Perhaps we should add //go:nosfloat to make
the compiler enforce this?)
Fixes #10641.
Change-Id: Ida4742c49000fae4fea4649f28afde630ce4c576 Signed-off-by: Shenghou Ma <minux@golang.org>
Reviewed-on: https://go-review.googlesource.com/9570 Reviewed-by: Dave Cheney <dave@cheney.net> Reviewed-by: Keith Randall <khr@golang.org>
Shenghou Ma [Sat, 2 May 2015 02:09:54 +0000 (22:09 -0400)]
go/build: reserve GOARCH values for all common architectures
Whenever we introduce a new GOARCH, older Go releases won't
recognize them and this causes trouble for both our users and
us (we need to add unnecessary build tags).
Go 1.5 has introduced three new GOARCHes so far: arm64 ppc64
ppc64le, we can take the time to introduce GOARCHes for all
common architectures that Go might support in the future to
avoid the problem.
Fixes #10165.
Change-Id: Ida4f9112897cfb1e85b06538db79125955ad0f4c
Reviewed-on: https://go-review.googlesource.com/9644 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Dave Cheney [Sat, 2 May 2015 02:44:49 +0000 (12:44 +1000)]
cmd/internal/ld: delete Biobuf
Update #10652
This proposal deletes cmd/internal/ld.Biobuf and replaces all uses with
cmd/internal/obj.Biobuf. As cmd/internal/ld already imported cmd/internal/obj
there are no additional dependencies created.
Notes:
- ld.Boffset included more checks, so it was merged into obj.Boffset
- obj.Bflush was removed in 8d16253c90ae, so replaced all calls to
ld.Bflush, with obj.Biobuf.Flush.
- Almost all of this change was prepared with sed.
Change-Id: I814854d52f5729a5a40c523c8188e465246b88da
Reviewed-on: https://go-review.googlesource.com/9660 Reviewed-by: Keith Randall <khr@golang.org>
Run-TryBot: Dave Cheney <dave@cheney.net>
Shenghou Ma [Sun, 3 May 2015 05:29:52 +0000 (01:29 -0400)]
go/internal/gcimporter, go/types: also skip tests on nacl/arm
Change-Id: I3e839587626832da069d95a7d7389ea6bb2318da
Reviewed-on: https://go-review.googlesource.com/9674 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Austin Clements [Mon, 4 May 2015 18:37:45 +0000 (14:37 -0400)]
runtime: add pointer size to type structure
This adds a field to the runtime type structure that records the size
of the prefix of objects of that type containing pointers. Any data
after this offset is scalar data.
This is necessary for shrinking the type bitmaps to 1 bit and will
help the garbage collector efficiently estimate the amount of heap
that needs to be scanned.
Rob Pike [Mon, 4 May 2015 18:28:51 +0000 (11:28 -0700)]
fmt: catch overflow in width and prec calculations
Fixes #10674.
Change-Id: If3fae3244d87aeaa70815f499105c264394aa7ad
Reviewed-on: https://go-review.googlesource.com/9657 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Rick Hudson [Tue, 28 Apr 2015 13:07:51 +0000 (09:07 -0400)]
runtime: Reduce calls to shouldtriggergc
shouldtriggergc is slightly expensive due to the call overhead
and the use of an atomic. This CL reduces the number of time
one checks if a GC should be done from one at each allocation
to once when a span is allocated. Since shouldtriggergc is an
important abstraction simply hand inlining it, along with its
atomic instruction would lose the abstraction.
At the end of lexInsideAction(), we return lexInsideAction: this is the default
behaviour when we are still parsing an action. But some switch branches return
lexInsideAction too.
So let's ensure code consistency by always reaching the end of the
lexInsideAction function when needed.
Change-Id: I7e9d8d6e51f29ecd6db6bdd63b36017845d95368
Reviewed-on: https://go-review.googlesource.com/9441 Reviewed-by: Rob Pike <r@golang.org>
Rob Pike [Fri, 1 May 2015 22:33:08 +0000 (15:33 -0700)]
text/template/parse: huge integers are not floats
Ideal constants in the template package are a little different from Go.
This is a case that slipped through the cracks: A huge integer number
was accepted as a floating-point number, but this loses precision
and is confusing. Also, the code in the template package (as opposed
to the parse package) wasn't expecting it.
Root this out at the source: If an integer doesn't fit an int64 or uint64,
complain right away.
Shenghou Ma [Sat, 2 May 2015 08:36:53 +0000 (04:36 -0400)]
cmd/internal/gc: fix build on big endian systems
The siz argument to both runtime.newproc and runtime.deferproc is
int32, not uintptr. This problem won't manifest on little-endian
systems because that stack slot is uintptr sized anyway. However,
on big-endian systems, it will make a difference.
Didier Spezia [Sat, 2 May 2015 11:03:35 +0000 (11:03 +0000)]
text/template: check for literals in chain of terms
The current parser ignores obvious errors such as:
{{0.1.E}}
{{true.any}}
{{"hello".wrong}}
{{nil.E}}
The common problem is that a chain is built from
a literal value. It then panics at execution time.
Furthermore, a double dot triggers the same behavior:
{{..E}}
Addresses a TODO left in Tree.operand to catch these
errors at parsing time.
Note that identifiers can include a '.', and pipelines
could return an object which a field can be derived
from (like a variable), so they are excluded from the check.
Fixes #10615
Change-Id: I903706d1c17861b5a8354632c291e73c9c0bc4e1
Reviewed-on: https://go-review.googlesource.com/9621 Reviewed-by: Rob Pike <r@golang.org>
Dave Cheney [Fri, 1 May 2015 05:29:11 +0000 (15:29 +1000)]
cmd/internal/obj: remove Biobuf unget
This change applies CL 9365 to the copy of Biobuf in cmd/internal/obj.
In the process I discovered that some of the methods that should have been
checking the unget buffer before reading were not and it was probably just
dumb luck that we handn't hit these issues before; Bungetc is only used in
one place in cmd/internal/gc and only an unlikely code path.
Change-Id: Ifa0c5c08442e9fe951a5078c6e9ec77a8a4dc2ff
Reviewed-on: https://go-review.googlesource.com/9529 Reviewed-by: Daniel Morsing <daniel.morsing@gmail.com> Reviewed-by: Ian Lance Taylor <iant@golang.org> Reviewed-by: Dave Cheney <dave@cheney.net>
Run-TryBot: Dave Cheney <dave@cheney.net>
Russ Cox [Fri, 1 May 2015 15:52:56 +0000 (11:52 -0400)]
runtime: correct accounting of scan work and bytes marked
(1) Count pointer-free objects found during scanning roots
as marked bytes, by not zeroing the mark total after scanning roots.
(2) Don't count the bytes for the roots themselves, by not adding
them to the mark total in scanblock (the zeroing removed by (1)
was aimed at that add but hitting more).
Combined, (1) and (2) fix the calculation of the marked heap size.
This makes the GC trigger much less often in the Go 1 benchmarks,
which have a global []byte pointing at 256 MB of data.
That 256 MB allocation was not being included in the heap size
in the current code, but was included in Go 1.4.
This is the source of much of the relative slowdown in that directory.
(3) Count the bytes for the roots as scanned work, by not zeroing
the scan total after scanning roots. There is no strict justification
for this, and it probably doesn't matter much either way,
but it was always combined with another buggy zeroing
(removed in (1)), so guilty by association.
cmd/internal/gc, runtime: use 1-bit bitmap for stack frames, data, bss
The bitmaps were 2 bits per pointer because we needed to distinguish
scalar, pointer, multiword, and we used the leftover value to distinguish
uninitialized from scalar, even though the garbage collector (GC) didn't care.
Now that there are no multiword structures from the GC's point of view,
cut the bitmaps down to 1 bit per pointer, recording just live pointer vs not.
The GC assumes the same layout for stack frames and for the maps
describing the global data and bss sections, so change them all in one CL.
The code still refers to 4-bit heap bitmaps and 2-bit "type bitmaps", since
the 2-bit representation lives (at least for now) in some of the reflect data.
Because these stack frame bitmaps are stored directly in the rodata in
the binary, this CL reduces the size of the 6g binary by about 1.1%.
Performance change is basically a wash, but using less memory,
and smaller binaries, and enables other bitmap reductions.
David Chase [Fri, 1 May 2015 15:16:35 +0000 (11:16 -0400)]
cmd/internal/gc: Toughen escape analysis against some bugs.
Ensures that parameter flow bits are not set for tags EscScope, EscHeap, EscNever;
crash the compiler earl to expose faulty logic, rather than flake out silently downstream.
David Chase [Thu, 26 Mar 2015 20:36:15 +0000 (16:36 -0400)]
cmd/internal/gc: improve flow of input params to output params
This includes the following information in the per-function summary:
outK = paramJ encoded in outK bits for paramJ
outK = *paramJ encoded in outK bits for paramJ
heap = paramJ EscHeap
heap = *paramJ EscContentEscapes
Note that (currently) if the address of a parameter is taken and
returned, necessarily a heap allocation occurred to contain that
reference, and the heap can never refer to stack, therefore the
parameter and everything downstream from it escapes to the heap.
The per-function summary information now has a tuneable number of bits
(2 is probably noticeably better than 1, 3 is likely overkill, but it
is now easy to check and the -m debugging output includes information
that allows you to figure out if more would be better.)
A new test was added to check pointer flow through struct-typed and
*struct-typed parameters and returns; some of these are sensitive to
the number of summary bits, and ought to yield better results with a
more competent escape analysis algorithm. Another new test checks
(some) correctness with array parameters, results, and operations.
The old analysis inferred a piece of plan9 runtime was non-escaping by
counteracting overconservative analysis with buggy analysis; with the
bug fixed, the result was too conservative (and it's not easy to fix
in this framework) so the source code was tweaked to get the desired
result. A test was added against the discovered bug.
The escape analysis was further improved splitting the "level" into
3 parts, one tracking the conventional "level" and the other two
computing the highest-level-suffix-from-copy, which is used to
generally model the cancelling effect of indirection applied to
address-of.
With the improved escape analysis enabled, it was necessary to
modify one of the runtime tests because it now attempts to allocate
too much on the (small, fixed-size) G0 (system) stack and this
failed the test.
Compiling src/std after touching src/runtime/*.go with -m logging
turned on shows 420 fewer heap allocation sites (10538 vs 10968).
Profiling allocations in src/html/template with
for i in {1..5} ;
do go tool 6g -memprofile=mastx.${i}.prof -memprofilerate=1 *.go;
go tool pprof -alloc_objects -text mastx.${i}.prof ;
done
showed a 15% reduction in allocations performed by the compiler.
David Crawshaw [Thu, 30 Apr 2015 20:38:10 +0000 (16:38 -0400)]
runtime/cgo, cmd/dist: turn off exc_bad_access handler by default
App Store policy requires programs do not reference the exc_server
symbol. (Some public forum threads show that Unity ran into this
several years ago and it is a hard policy rule.) While some research
suggests that I could write my own version of exc_server, the
expedient course is to disable the exception handler by default.
Go programs only need it when running under lldb, which is primarily
used by tests. So enable the exception handler in cmd/dist when we
are running the tests.
Dave Cheney [Fri, 1 May 2015 00:15:33 +0000 (10:15 +1000)]
cmd/8g: don't call gc.Fatal during initalisation
Fixes #10592
Calling gc.Fatal before gc.Main has been called ends up flushing gc.bstdout before
it is properly set up. Ideally obj.Bflush would handle this case, but that type
and its callers are rather convoluted, so take the simpler route and avoid calling
gc.Fatal altogether.
Change-Id: I338b469e86edba558b6bedff35bb904bfc3d6990
Reviewed-on: https://go-review.googlesource.com/9525 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>
Rob Pike [Thu, 30 Apr 2015 21:37:43 +0000 (14:37 -0700)]
cmd/cover: copy to standard repository from golang.org/x/tools/cmd/cover
This required dealing with the ill-advised split of the profile code
into a separate package. I just copied it over unchanged. The package
does not deserve to be in the standard repository. We can cope
with the duplication.
Also update the go command to know about the new location.
Michael Hudson-Doyle [Wed, 29 Apr 2015 10:58:52 +0000 (22:58 +1200)]
cmd/internal/ld: put the list of packages built into a shared library into an ELF note
Change-Id: I611f7dec2109dc7e2f090ced0a1dca3d4b577134
Reviewed-on: https://go-review.googlesource.com/9520 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Change-Id: I5ca7366ba8bf6221a36d25a2157dda4b4f3e16fa
Reviewed-on: https://go-review.googlesource.com/9523 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Ian Lance Taylor [Wed, 29 Apr 2015 21:32:48 +0000 (14:32 -0700)]
cmd/go, cmd/cgo: support -buildmode=c-archive for gccgo
This extends the cgo changes in http://golang.org/cl/8094 to gccgo.
It also adds support for setting runtime_iscgo correctly for gccgo;
the gc runtime bases the variable on the runtime/cgo package, but
gccgo has no equivalent to that package.
The go tool supports -buildmode=c-archive for gccgo by linking all the
Go objects together using -r. For convenience this object is then put
into an archive file.
The go tool now passes -fsplit-stack when building C code for gccgo on
386 and amd64. This is required for using -r and will also cut down
on unnecessary stack splits.
The go tool no longer applies standard package cgo LDFLAGS when using
gccgo. This is mainly to avoid getting confused by the LDFLAGS in the
runtime/cgo package that gccgo does not use.
Change-Id: I1d0865b2a362818a033ca9e9e901d0ce250784e7
Reviewed-on: https://go-review.googlesource.com/9511 Reviewed-by: David Crawshaw <crawshaw@golang.org>
html/template: fix quadratic performance with special tags
The current implementation of the tSpecialTagEnd function
is inefficient since it generates plenty of memory allocations
and converts the whole buffer to lowercase at each call.
If the number of special tags increases linearly with the
template size, the complexity becomes quadratic.
This CL provides an alternative implementation.
While the algorithm is probably still not optimal, it avoids
the quadratic behavior and the memory allocations.
benchmark old ns/op new ns/op delta
BenchmarkTemplateSpecialTags-4 19326431 532190 -97.25%
benchmark old allocs new allocs delta
BenchmarkTemplateSpecialTags-4 2650 190 -92.83%
benchmark old bytes new bytes delta
BenchmarkTemplateSpecialTags-4 4106460 46568 -98.87%
While we are there, make sure we respect the HTML tokenization algorithm.
An end tag needs to be followed by a space, tab, CR, FF, /, or > as described
in https://html.spec.whatwg.org/multipage/syntax.html#tokenization
Explicitly add this check.
Fixes #10605
Change-Id: Ia33ddee164ab608a69ac4183e16ec506bbeaa54c
Reviewed-on: https://go-review.googlesource.com/9502 Reviewed-by: Rob Pike <r@golang.org>
Schedule the work as early as possible, while still respecting the
utilization percentage on average. The old code tried never to
go above the utilization percentage. The new code is willing
to go above the utilization percentage by one time slice
(but of course after doing that it must wait until the percentage
drops back down to the target before it gets another time slice).
The effect is that for concurrent GCs that can run in a small number
of time slices, the time during which write barriers are enabled is
reduced by one mutator + GC time slice round (possibly 30 ms per GC).
This only affects the fractional GC processor (the remainder of GOMAXPROCS/4),
so it matters most in GOMAXPROCS=1, a bit in GOMAXPROCS=2, and not at
all in GOMAXPROCS=4.
gcDumpObject is used to print the source and destination objects when
checkmark find a missing mark. However, gcDumpObject currently assumes
the given pointer will point to a heap object. This is not true of the
source object during root marking and may not even be true of the
destination object in the limited situations where the heap points
back in to the stack.
If the pointer isn't a heap object, gcDumpObject will attempt an
out-of-bounds access to h_spans. This will cause a panicslice, which
will attempt to construct a useful panic message. This will cause a
string allocation, which will lead mallocgc to panic because the GC is
in mark termination (checkmark only happens during mark termination).
Fix this by checking that the pointer points into the heap arena
before attempting to use it as an arena pointer.
Change-Id: I09da600c380d4773f1f8f38e45b82cb229ea6382
Reviewed-on: https://go-review.googlesource.com/9498 Reviewed-by: Rick Hudson <rlh@golang.org>
Currently the packages have the following index functions:
func Index(s, sep []byte) int
func IndexAny(s []byte, chars string) int
func IndexByte(s []byte, c byte) int
func IndexFunc(s []byte, f func(r rune) bool) int
func IndexRune(s []byte, r rune) int
func LastIndex(s, sep []byte) int
func LastIndexAny(s []byte, chars string) int
func LastIndexFunc(s []byte, f func(r rune) bool) int
Searching for the last occurrence of a byte is quite common
for string parsing algorithms (e.g. find the last paren on a line).
Also addition of LastIndexByte makes the set more orthogonal.
Change-Id: Ida168849acacf8e78dd70c1354bef9eac5effafe
Reviewed-on: https://go-review.googlesource.com/9500 Reviewed-by: Rob Pike <r@golang.org>
Alex Brainman [Thu, 23 Apr 2015 05:04:35 +0000 (15:04 +1000)]
mime, time, internal/syscall/windows/registry: use new registry package to simplify code
This CL copies golang.org/x/sys/windows/registry into
internal/syscall/windows/registry (minus KeyInfo.ModTime to prevent
dependency cycles). New registry package is used in mime and time
packages instead of calling Windows API directly.
Bryan Ford [Fri, 19 Dec 2014 19:28:44 +0000 (14:28 -0500)]
math/big: add modular square-root and Jacobi functions
This change adds Int.ModSqrt to compute modular square-roots via the
standard Tonelli-Shanks algorithm, and the Jacobi function that this and
many other modular-arithmetic algorithms depend on.
This is needed by change 1883 (https://golang.org/cl/1883), to add
support for ANSI-standard compressed encoding of elliptic curve points.
Change-Id: Icc4805001bba0b3cb7200e0b0a7f87b14a9e9439
Reviewed-on: https://go-review.googlesource.com/1886 Reviewed-by: Adam Langley <agl@golang.org>
Adam Langley [Wed, 29 Apr 2015 17:36:38 +0000 (10:36 -0700)]
crypto/x509: be strict about trailing data.
The X.509 parser was allowing trailing data after a number of structures
in certificates and public keys. There's no obvious security issue here,
esp in certificates which are signed anyway, but this change makes
trailing data an error just in case.
Adam Langley [Wed, 29 Apr 2015 17:10:48 +0000 (10:10 -0700)]
crypto/tls: update the supported signature algorithms.
This is the second in a two-part change. See https://golang.org/cl/9415
for details of the overall change.
This change updates the supported signature algorithms to include
SHA-384 and updates all the testdata/ files accordingly. Even some of
the testdata/ files named “TLS1.0” and “TLS1.1” have been updated
because they have TLS 1.2 ClientHello's even though the server picks a
lower version.
Adam Langley [Tue, 28 Apr 2015 16:13:38 +0000 (09:13 -0700)]
crypto/tls: decouple handshake signatures from the handshake hash.
Prior to TLS 1.2, the handshake had a pleasing property that one could
incrementally hash it and, from that, get the needed hashes for both
the CertificateVerify and Finished messages.
TLS 1.2 introduced negotiation for the signature and hash and it became
possible for the handshake hash to be, say, SHA-384, but for the
CertificateVerify to sign the handshake with SHA-1. The problem is that
one doesn't know in advance which hashes will be needed and thus the
handshake needs to be buffered.
Go ignored this, always kept a single handshake hash, and any signatures
over the handshake had to use that hash.
However, there are a set of servers that inspect the client's offered
signature hash functions and will abort the handshake if one of the
server's certificates is signed with a hash function outside of that
set. https://robertsspaceindustries.com/ is an example of such a server.
Clearly not a lot of thought happened when that server code was written,
but its out there and we have to deal with it.
This change decouples the handshake hash from the CertificateVerify
hash. This lays the groundwork for advertising support for SHA-384 but
doesn't actually make that change in the interests of reviewability.
Updating the advertised hash functions will cause changes in many of the
testdata/ files and some errors might get lost in the noise. This change
only needs to update four testdata/ files: one because a SHA-384-based
handshake is now being signed with SHA-256 and the others because the
TLS 1.2 CertificateRequest message now includes SHA-1.
This change also has the effect of adding support for
client-certificates in SSLv3 servers. However, SSLv3 is now disabled by
default so this should be moot.
It would be possible to avoid much of this change and just support
SHA-384 for the ServerKeyExchange as the SKX only signs over the nonces
and SKX params (a design mistake in TLS). However, that would leave Go
in the odd situation where it advertised support for SHA-384, but would
only use the handshake hash when signing client certificates. I fear
that'll just cause problems in the future.
Much of this code was written by davidben@ for the purposes of testing
BoringSSL.
Partly addresses #9757
Change-Id: I5137a472b6076812af387a5a69fc62c7373cd485
Reviewed-on: https://go-review.googlesource.com/9415
Run-TryBot: Adam Langley <agl@golang.org> Reviewed-by: Adam Langley <agl@golang.org>
Change-Id: I6cebaf42f2596c7f8fef3a67afb1e5ccb428d09c
Reviewed-on: https://go-review.googlesource.com/9521 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Michael Hudson-Doyle [Mon, 27 Apr 2015 03:00:48 +0000 (15:00 +1200)]
misc/cgo/testshared: add basic test for -buildmode=shared/-linkshared
Just a first basic test, I'll extend this to test more but want to get an
opinion on basic approach first.
Change-Id: Idab9ebd7d9960b000b81a01a1e53258bf4bce755
Reviewed-on: https://go-review.googlesource.com/9386 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
This change deflakes timeout, deadline tests, especially fixes socket
and goroutine leaks. Also adds a few missing tests that use features
introduced after go1 release.
Change-Id: Ibf73a4859f8d4a0ee494ca2fd180cbce72a7a2c7
Reviewed-on: https://go-review.googlesource.com/9464 Reviewed-by: Ian Lance Taylor <iant@golang.org>
This change merges unicast_posix_test.go and multicast_test.go into
listen_test.go before deflaking tests for Listen functions.
No code changes.
Change-Id: Ic4cd6531b95dfb5b6e6e254241692eca61a71e94
Reviewed-on: https://go-review.googlesource.com/9460 Reviewed-by: Ian Lance Taylor <iant@golang.org>
This change adds missing CloseRead test and Close tests on Conn,
Listener and PacketConn with various networks.
Change-Id: Iadf99eaf526a323f853d203edc7c8d0577f67972
Reviewed-on: https://go-review.googlesource.com/9469 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Peter Waldschmidt [Wed, 29 Apr 2015 20:59:38 +0000 (16:59 -0400)]
net/http: Don't set Content-Length: -1 when responding to a POST
Fixes an issue where Response.Write writes out a Content-Length: -1
header when the corresponding Request is a POST or PUT and the
ContentLength was not previously set.
This was encountered when using httputil.DumpResponse
to write out the response from a server that responded to a PUT
request with no Content-Length header. The dumped output is
thus invalid.
Not only by network, transport-layer intermediaries but by
virtualization stuff in a node, it is hard to identify the root cause of
weird faults without information of packet flows after disaster
happened.
This change adds Source field to OpError to be able to represent a
5-tuple of internet transport protocols for helping dealing with
complicated systems.
Also clarifies the usage of Source and Addr fields.
Updates #4856.
Change-Id: I96a523fe391ed14406bfb21604c461d4aac2fa19
Reviewed-on: https://go-review.googlesource.com/9231 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Change-Id: I5569dcdefe8adba346810124b16721674956bce6
Reviewed-on: https://go-review.googlesource.com/9515 Reviewed-by: Ian Lance Taylor <iant@golang.org>
net: allow a dns TXT record to contain more than one <character-string>
RFC 1035 3.3.14 allows a TXT record to contain one or more <character-string>s.
The current implementation returns a "no such host" error if there is more
than one <character-string> in the TXT record.
Fixes #10482
Change-Id: I0ded258005e6b7ba45f687fecd10afa2b321bb77
Reviewed-on: https://go-review.googlesource.com/8966 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>