Emmanuel Odeke [Sun, 16 Oct 2016 12:00:27 +0000 (05:00 -0700)]
net/http: support multiple identical Content-Length headers
Referencing RFC 7230 Section 3.3.2, this CL
deduplicates multiple identical Content-Length headers
of a message or rejects the message as invalid if the
Content-Length values differ.
Daniel Theophanes [Mon, 3 Oct 2016 16:49:25 +0000 (09:49 -0700)]
database/sql: add option to use named parameter in query arguments
Modify the new Context methods to take a name-value driver struct.
This will require more modifications to drivers to use, but will
reduce the overall number of structures that need to be maintained
over time.
Brad Fitzpatrick [Sun, 16 Oct 2016 09:27:29 +0000 (10:27 +0100)]
net/http: add more docs on ErrHijacked
Updates #16456
Change-Id: Ifea651ea3ece2267a6f0c1638181d6ddd9248a9f
Reviewed-on: https://go-review.googlesource.com/31181 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Brad Fitzpatrick [Sat, 15 Oct 2016 15:56:51 +0000 (16:56 +0100)]
syscall, net: make deadline changes affect blocked read/write calls on nacl
Flesh out nacl's fake network system to match how all the other
platforms work: all other systems' SetReadDeadline and
SetWriteDeadline affect currently-blocked read & write calls.
This was documented in golang.org/cl/30164 because it was the status
quo and existing packages relied on it. (notably the net/http package)
And add a test.
Change-Id: I074a1054dcabcedc97b173dad5e827f8babf7cfc
Reviewed-on: https://go-review.googlesource.com/31178
Run-TryBot: Brad Fitzpatrick <bradfitz@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Caleb Spare [Fri, 14 Oct 2016 07:59:19 +0000 (00:59 -0700)]
html/template: fix Clone so that t.Lookup(t.Name()) yields t
Template.escape makes the assumption that t.Lookup(t.Name()) is t
(escapeTemplate looks up the associated template by name and sets
escapeErr appropriately).
This assumption did not hold for a Cloned template, because the template
associated with t.Name() was a second copy of the original.
Add a test for the assumption that t.Lookup(t.Name()) == t.
One effect of this broken assumption was #16101: parallel Executes
racily accessed the template namespace because each Execute call saw
t.escapeErr == nil and re-escaped the template concurrently with read
accesses occurring outside the namespace mutex.
Add a test for this race.
Related to #12996 and CL 16104.
Fixes #16101
Change-Id: I59831d0847abbabb4ef9135f2912c6ce982f9837
Reviewed-on: https://go-review.googlesource.com/31092
Run-TryBot: Rob Pike <r@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Rob Pike <r@golang.org>
Alex Brainman [Fri, 14 Oct 2016 06:03:01 +0000 (17:03 +1100)]
misc/cgo/testcarchive: do not use same executable name in TestInstall
Fixes #17439
Change-Id: I7caa28519f38692f9ca306f0789cbb975fa1d7c4
Reviewed-on: https://go-review.googlesource.com/31112 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Alex Brainman <alex.brainman@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Rob Pike [Sun, 16 Oct 2016 18:28:45 +0000 (11:28 -0700)]
testing: mention in docs for Logf that a final newline is added if needed
Fixes #16423
Change-Id: I9635db295be4d356d427adadd309084e16c4582f
Reviewed-on: https://go-review.googlesource.com/31255 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Austin Clements [Fri, 14 Oct 2016 21:35:18 +0000 (17:35 -0400)]
test: simplify fixedbugs/issue15747.go
The error check patterns in this test are more complex than necessary
because f2 gets inlined into f1. This behavior isn't important to the
test, so disable inlining of f2 and simplify the error check patterns.
Austin Clements [Mon, 10 Oct 2016 21:14:14 +0000 (17:14 -0400)]
runtime: use more go:nowritebarrierrec in proc.go
Currently we use go:nowritebarrier in many places in proc.go.
go:notinheap and go:yeswritebarrierrec now let us use
go:nowritebarrierrec (the recursive form of the go:nowritebarrier
pragma) more liberally. Do so in proc.go
Change-Id: Ia7fcbc12ce6c51cb24730bf835fb7634ad53462f
Reviewed-on: https://go-review.googlesource.com/30942 Reviewed-by: Rick Hudson <rlh@golang.org>
Austin Clements [Tue, 11 Oct 2016 15:47:14 +0000 (11:47 -0400)]
runtime: make mSpanList more go:notinheap-friendly
Currently mspan links to its previous mspan using a **mspan field that
points to the previous span's next field. This simplifies some of the
list manipulation code, but is going to make it very hard to convince
the compiler that mspan list manipulations don't need write barriers.
Fix this by using a more traditional ("boring") linked list that uses
a simple *mspan pointer to the previous mspan. This complicates some
of the list manipulation slightly, but it will let us eliminate all
write barriers from the mspan list manipulation code by marking mspan
go:notinheap.
Change-Id: I0d0b212db5f20002435d2a0ed2efc8aa0364b905
Reviewed-on: https://go-review.googlesource.com/30940 Reviewed-by: Rick Hudson <rlh@golang.org>
Austin Clements [Wed, 12 Oct 2016 02:53:27 +0000 (22:53 -0400)]
cmd/compile: add go:notinheap type pragma
This adds a //go:notinheap pragma for declarations of types that must
not be heap allocated. We ensure these rules by disallowing new(T),
make([]T), append([]T), or implicit allocation of T, by disallowing
conversions to notinheap types, and by propagating notinheap to any
struct or array that contains notinheap elements.
The utility of this pragma is that we can eliminate write barriers for
writes to pointers to go:notinheap types, since the write barrier is
guaranteed to be a no-op. This will let us mark several scheduler and
memory allocator structures as go:notinheap, which will let us
disallow write barriers in the scheduler and memory allocator much
more thoroughly and also eliminate some problematic hybrid write
barriers.
This also makes go:nowritebarrierrec and go:yeswritebarrierrec much
more powerful. Currently we use go:nowritebarrier all over the place,
but it's almost never what you actually want: when write barriers are
illegal, they're typically illegal for a whole dynamic scope. Partly
this is because go:nowritebarrier has been around longer, but it's
also because go:nowritebarrierrec couldn't be used in situations that
had no-op write barriers or where some nested scope did allow write
barriers. go:notinheap eliminates many no-op write barriers and
go:yeswritebarrierrec makes it possible to opt back in to write
barriers, so these two changes will let us use go:nowritebarrierrec
far more liberally.
This updates #13386, which is about controlling pointers from non-GC'd
memory to GC'd memory. That would require some additional pragma (or
pragmas), but could build on this pragma.
Change-Id: I6314f8f4181535dd166887c9ec239977b54940bd
Reviewed-on: https://go-review.googlesource.com/30939 Reviewed-by: Keith Randall <khr@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
This pragma cancels the effect of go:nowritebarrierrec. This is useful
in the scheduler because there are places where we enter a function
without a valid P (and hence cannot have write barriers), but then
obtain a P. This allows us to annotate the function with
go:nowritebarrierrec and split out the part after we've obtained a P
into a go:yeswritebarrierrec function.
Change-Id: Ic8ce4b6d3c074a1ecd8280ad90eaf39f0ffbcc2a
Reviewed-on: https://go-review.googlesource.com/30938 Reviewed-by: Matthew Dempsky <mdempsky@google.com> Reviewed-by: Keith Randall <khr@golang.org>
Daniel Theophanes [Thu, 6 Oct 2016 18:06:21 +0000 (11:06 -0700)]
database/sql: add support for multiple result sets
Many database systems allow returning multiple result sets
in a single query. This can be useful when dealing with many
intermediate results on the server and there is a need
to return more then one arity of data to the client.
Rob Pike [Thu, 13 Oct 2016 02:51:02 +0000 (19:51 -0700)]
cmd/cover: handle gotos
If a labeled statement is the target of a goto, we must treat it as the
boundary of a new basic block, but only if it is not already the boundary
of a basic block such as a labeled for loop.
Fixes #16624
Now reports 100% coverage for the test in the issue.
Change-Id: If118bb6ff53a96c738e169d92c03cb3ce97bad0e
Reviewed-on: https://go-review.googlesource.com/30977 Reviewed-by: Ian Lance Taylor <iant@golang.org> Reviewed-by: Robert Griesemer <gri@golang.org>
Anthony Canino [Sat, 10 Oct 2015 19:24:34 +0000 (15:24 -0400)]
cmd/compile: "abc"[1] is not an ideal constant
"abc"[1] is not like 'b', in that -"abc"[1] is uint8 math, not ideal constant math.
Delay the constantification until after ideal constant folding is over.
Ian Lance Taylor [Thu, 13 Oct 2016 00:09:54 +0000 (17:09 -0700)]
cmd/cgo: use alias for unsafe rather than separate functions
When we need to generate a call to _cgoCheckPointer, we need to type
assert the result back to the desired type. That is harder when the type
is unsafe.Pointer, as the package can have values of unsafe.Pointer
types without actually importing unsafe, by mixing C void* and :=. We
used to handle this by generating a special function for each needed
type, and defining that function in a separate file where we did import
unsafe.
Simplify the code by not generating those functions, but instead just
import unsafe under the alias _cgo_unsafe. This is a simplification step
toward a fix for issue #16591.
Change-Id: I0edb3e04b6400ca068751709fe063397cf960a54
Reviewed-on: https://go-review.googlesource.com/30973
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Matthew Dempsky [Wed, 12 Oct 2016 22:48:18 +0000 (15:48 -0700)]
cmd/compile: add OSTRUCTKEY for keyed struct literals
Previously, we used OKEY nodes to represent keyed struct literal
elements. The field names were represented by an ONAME node, but this
is clumsy because it's the only remaining case where ONAME was used to
represent a bare identifier and not a variable.
This CL introduces a new OSTRUCTKEY node op for use in struct
literals. These ops instead store the field name in the node's own Sym
field. This is similar in spirit to golang.org/cl/20890.
Significant reduction in allocations for struct literal heavy code
like package unicode:
Alex Brainman [Wed, 21 Sep 2016 01:19:36 +0000 (11:19 +1000)]
os: make readConsole handle its input and output correctly
This CL introduces first test for readConsole. And new test
discovered couple of problems with readConsole.
Console characters consist of multiple bytes each, but byte blocks
returned by syscall.ReadFile have no character boundaries. Some
multi-byte characters might start at the end of one block, and end
at the start of next block. readConsole feeds these blocks to
syscall.MultiByteToWideChar to convert them into utf16, but if some
multi-byte characters have no ending or starting bytes, the
syscall.MultiByteToWideChar might get confused. Current version of
syscall.MultiByteToWideChar call will make
syscall.MultiByteToWideChar ignore all these not complete
multi-byte characters.
The CL solves this issue by changing processing from "randomly
sized block of bytes at a time" to "one multi-byte character at a
time". New readConsole code calls syscall.ReadFile to get 1 byte
first. Then it feeds this byte to syscall.MultiByteToWideChar.
The new syscall.MultiByteToWideChar call uses MB_ERR_INVALID_CHARS
flag to make syscall.MultiByteToWideChar return error if input is
not complete character. If syscall.MultiByteToWideChar returns
correspondent error, we read another byte and pass 2 byte buffer
into syscall.MultiByteToWideChar, and so on until success.
Old readConsole code would also sometimes return no data if user
buffer was smaller then uint16 size, which would confuse callers
that supply 1 byte buffer. This CL fixes that problem too.
Hiroshi Ioka [Wed, 12 Oct 2016 02:34:47 +0000 (11:34 +0900)]
cmd/compile/internal/gc: cleanup esc.go
* convert important functions to methods
* rename EscXXX to XXX in NodeEscState
* rename local variables more friendly
* simplify redundant code
* update comments
Change-Id: I8442bf4f8dde84523d9a2ad3d04b1cd326bd4719
Reviewed-on: https://go-review.googlesource.com/30893
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: David Chase <drchase@google.com>
Alex Browne [Sun, 8 Nov 2015 04:54:41 +0000 (23:54 -0500)]
cmd/vet: check for duplicate json, xml struct field tags
It is easy to make the mistake of duplicating json struct field
tags especially when copy/pasting. This commit causes go vet to
report the mistake. Only field tags in the same struct type are
considered, because that is the only case which is undoubtedly an
error.
Allan Simon [Sat, 10 Oct 2015 20:16:58 +0000 (04:16 +0800)]
encoding/xml: prevent omitempty from omitting non-nil pointers to empty values
There was an inconsistency between the (json encoding + documentation)
and the xml encoding implementation. Pointer to an empty value was
not being serialized (i.e simply ignored). Which had the effect of making
impossible to have a struct with a string field for which we wanted to
serialize the value ""
Lynn Boger [Wed, 12 Oct 2016 12:50:30 +0000 (07:50 -0500)]
cmd/asm: recognize CR1-CR7 on ppc64x branch instructions
Some of the branch instructions (BEQ, BNE, BLT, etc.) accept
all the valid CR values as operands, but the CR register value is
not parsed and not put into the instruction, so that CR0 is always
used regardless of what was specified on the instruction. For example
BEQ CR2,label becomes beq cr0,label.
This adds the change to the PPC64 assembler to recognize the CR value
and set the approppriate field in the instruction so the correct
CR is used. This also adds some general comments on the branch
instruction BC and its operand values.
Keith Randall [Tue, 11 Oct 2016 15:36:38 +0000 (08:36 -0700)]
cmd/compile,runtime: redo how map assignments work
To compile:
m[k] = v
instead of:
mapassign(maptype, m, &k, &v), do
do:
*mapassign(maptype, m, &k) = v
mapassign returns a pointer to the value slot in the map. It is just
like mapaccess except that it will allocate a new slot if k is not
already present in the map.
This makes map accesses faster but potentially larger (codewise).
It is faster because the write into the map is done when the compiler
knows the concrete type, so it can be done with a few store
instructions instead of calling typedmemmove. We also potentially
avoid stack temporaries to hold v.
The code can be larger when the map has pointers in its value type,
since there is a write barrier call in addition to the mapassign call.
That makes the code at the callsite a bit bigger (go binary is 0.3%
bigger).
This CL is in preparation for doing operations like m[k] += v with
only a single runtime call. That will roughly double the speed of
such operations.
Update #17133
Update #5147
Change-Id: Ia435f032090a2ed905dac9234e693972fe8c2dc5
Reviewed-on: https://go-review.googlesource.com/30815
Run-TryBot: Keith Randall <khr@golang.org> Reviewed-by: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Adam Langley [Tue, 11 Oct 2016 01:23:37 +0000 (18:23 -0700)]
crypto/tls: support X25519.
X25519 (RFC 7748) is now commonly used for key agreement in TLS
connections, as specified in
https://tools.ietf.org/html/draft-ietf-tls-curve25519-01.
This change adds support for that in crypto/tls, but does not enabled it
by default so that there's less test noise. A future change will enable
it by default and will update all the test data at the same time.
Adam Langley [Tue, 11 Oct 2016 17:08:57 +0000 (10:08 -0700)]
crypto/tls: switch to OpenSSL 1.1.0 for test data.
We will need OpenSSL 1.1.0 in order to test some of the features
expected for Go 1.8. However, 1.1.0 also disables (by default) some
things that we still want to test, such as RC4, 3DES and SSLv3. Thus
developers wanting to update the crypto/tls test data will need to build
OpenSSL from source.
This change updates the test data with transcripts generated by 1.1.0
(in order to reduce future diffs) and also causes a banner to be printed
if 1.1.0 is not used when updating.
(The test for an ALPN mismatch is removed because OpenSSL now terminates
the connection with a fatal alert if no known ALPN protocols are
offered. There's no point testing against this because it's an OpenSSL
behaviour.)
Joe Tsai [Fri, 2 Sep 2016 23:17:37 +0000 (16:17 -0700)]
archive/tar: fix and cleanup readOldGNUSparseMap
* Assert that the format is GNU.
Both GNU and STAR have some form of sparse file support with
incompatible header structures. Worse yet, both formats use the
'S' type flag to indicate the presence of a sparse file.
As such, we should check the format (based on magic numbers)
and fail early.
* Move realsize parsing logic into readOldGNUSparseMap.
This is related to the sparse parsing logic and belongs here.
* Fix the termination condition for parsing sparse fields.
The termination condition for reading the sparse fields
is to simply check if the first byte of the offset field is NULL.
This does not seem to be documented in the GNU manual, but this is
the check done by the both the GNU and BSD implementations:
http://git.savannah.gnu.org/cgit/tar.git/tree/src/sparse.c?id=9a33077a7b7ad7d32815a21dee54eba63b38a81c#n731
https://github.com/libarchive/libarchive/blob/1fa9c7bf90f0862036a99896b0501c381584451a/libarchive/archive_read_support_format_tar.c#L2207
* Fix the parsing of sparse fields to use parseNumeric.
This is what GNU and BSD do. The previous two links show that
GNU and BSD both handle base-256 and base-8.
* Detect truncated streams.
The call to io.ReadFull does not check if the error is io.EOF.
Getting io.EOF in this method is never okay and should always be
converted to io.ErrUnexpectedEOF.
* Simplify the function.
The logic is essentially a do-while loop so we can remove
some redundant code.
Joe Tsai [Sat, 3 Sep 2016 04:03:57 +0000 (21:03 -0700)]
archive/tar: handle integer overflow on 32bit machines
Most calls to strconv.ParseInt(x, 10, 0) should really be
calls to strconv.ParseInt(x, 10, 64) in order to ensure that they
do not overflow on 32b architectures.
Furthermore, we should document a bug where Uid and Gid may
overflow on 32b machines since the type is declared as int.
- trim blocks with multiple predecessors
- trim blocks, which contain only phi-functions
- trim blocks, which can be merged into the successor block
As an example, compiling the following source:
---8<------
package p
type Node struct {
Key int
Left, Right *Node
}
func Search(r *Node, k int) *Node {
for r != nil {
switch {
case k == r.Key:
return r
case k < r.Key:
r = r.Left
default:
r = r.Right
}
}
return nil
}
---8<------
with `GOSSAFUNC=Search" go tool compile t.go`, results in the following
code:
The jump at 16 corresponds to the edge b21 -> b4. The block b4 contains
only phi-ops, i.e. no actual code besides the jump to b2. However b4 is
not trimmed, because it a) has more than one predecessor, and b) it is
not empty.
This change enhances trim.go to remove more blocks, subject to the
following criteria:
- block has predecessors (i.e. not the start block)
- block is BlockPlain
- block does not loop back to itself
- block is the single predecessor of its successor; the instructions of
the block are merged into the successor
- block does no emit actual code, besides a possible unconditional
jump.
Currently only OpPhi are considered to not be actual code,
perhaps OpKeepAlive/others should be considered too?
As an example, after the change, the block b4 is trimmed and the jump at
18 jumps directly to 8.
Revision 1: Adjust phi-ops arguments after merge
Ensure the number of phi-ops arguments matches the new number of
predecessors in the merged block.
When moving values, make them refer to the merged block.
Revision 2:
- Make clear the intent that we do not want to trim the entry block
- Double check that we are merging a phi operation
- Minor code style fix
- Fix a potentially dangerous situation when a blocks refers to the
inline value space in another block
Cyrill Schumacher [Sat, 20 Aug 2016 09:32:32 +0000 (11:32 +0200)]
net/http: optimize internal cookie functions
- precalculate *Cookie slice in read cookie functions
- readSetCookies: pre-allocs depending on the count of Set-Cookies
- rename success variable to ok; avoid else
- refactor Cookie.String to use less allocations
- remove fmt package and replace with writes to a bytes.Buffer
- add BenchmarkReadSetCookies and BenchmarkReadCookies
Joshua Boelter [Wed, 10 Aug 2016 05:37:26 +0000 (22:37 -0700)]
runtime: check for errors returned by windows sema calls
Add checks for failure of CreateEvent, SetEvent or
WaitForSingleObject. Any failures are considered fatal and
will throw() after printing an informative message.
David du Colombier [Wed, 12 Oct 2016 12:57:19 +0000 (14:57 +0200)]
cmd/link: use HEADR to define FlagTextAddr (cosmetic change)
This cosmetic change defines ld.FlagTextAddr using ld.HEADR in
the Plan 9 cases, like it is done for other operating systems.
Change-Id: Ic929c1c437f25661058682cf3e159f0b16cdc538
Reviewed-on: https://go-review.googlesource.com/30912
Run-TryBot: David du Colombier <0intro@gmail.com>
TryBot-Result: Gobot Gobot <gobot@golang.org> Reviewed-by: Ian Lance Taylor <iant@golang.org>
David du Colombier [Wed, 12 Oct 2016 12:24:39 +0000 (14:24 +0200)]
cmd/link: fix build on plan9/amd64
Support for multiple text sections was added in CL 27790.
However, this change broke the build on plan9/amd64.
In relocsym, the R_ADDROFF relocation was changed to
use offsets relative to the start of the first text
section. However, Segtext.Vaddr is the address of
the text segment, while we expect to start from
the first section (text.runtime) of the text segment.
Fixes #17411.
Change-Id: I86bbcbda81cea735b0ecf156eab2e6e5d63acce3
Reviewed-on: https://go-review.googlesource.com/30911
Run-TryBot: David du Colombier <0intro@gmail.com> Reviewed-by: Ian Lance Taylor <iant@golang.org>
Ian Lance Taylor [Wed, 12 Oct 2016 04:04:16 +0000 (21:04 -0700)]
syscall: unify NsecToTime{spec,val}, fix for times < 1970
All the implementations of NsecToTimespec and NsecToTimeval were the
same other than types. Write a single version that uses
GOARCH/GOOS-specific setTimespec and setTimeval functions to handle the
types.
The logic in NsecToTimespec and NsecToTimeval caused times before 1970
to have a negative usec/nsec. The Linux kernel requires that usec
contain a positive number; for consistency, we do this for both
NsecToTimespec and NsecToTimeval.
Dmitri Shuralyov [Tue, 26 Jul 2016 17:01:18 +0000 (13:01 -0400)]
path/filepath: remove unneeded doc statement for SplitList
This is a followup to CL 24747, where the package doc phrase
"Functions in this package replace occurrences of slash unless otherwise specified."
was removed. The phrase was originally added in CL 7310 together
with this explicit opt out statement for SplitList.
Remove it since it's no longer neccessary. This helps consistency.
Michael Pratt [Thu, 21 Jul 2016 05:01:33 +0000 (22:01 -0700)]
cmd/pprof: instruction-level granularity in callgrind output
When generating callgrind format output, produce cost lines at
instruction granularity. This allows visualizers supporting the
callgrind format to display instruction-level profiling information.
We also need to provide the object file (ob=) in order for tools to find
the object file to disassemble when displaying assembly.
We opportunistically group cost lines corressponding to the same
function together, reducing the number of superfluous description lines.
Subposition compression (relative position numbering) is also used to
reduce the output size.
Filippo Valsorda [Tue, 24 May 2016 11:50:38 +0000 (12:50 +0100)]
cmd/trace: add option to output pprof files
The trace tool can generate some interesting profiles, but it was only
exposing them as svg through the web UI. This adds command line options
to generate the raw pprof file.
James Clarke [Tue, 11 Oct 2016 18:08:45 +0000 (19:08 +0100)]
debug/elf: add sparc64 relocations
Change-Id: I1a2504ad9ca8607588d2d366598115fe360435b5
Reviewed-on: https://go-review.googlesource.com/30870 Reviewed-by: Ian Lance Taylor <iant@golang.org>
Joe Tsai [Tue, 11 Oct 2016 01:23:56 +0000 (18:23 -0700)]
archive/zip: only use Extended Timestamp on non-zero MS-DOS timestamps
We should preserve the fact that a roundtrip read on fields with the zero
value should remain the zero for those that are reasonable to stay that way.
If the zero value for a MS-DOS timestamp was used, then it is sensible for
that zero value to also be read back later.
Adam Langley [Mon, 10 Oct 2016 23:26:51 +0000 (16:26 -0700)]
crypto/x509: parse all names in an RDN.
The Subject and Issuer names in a certificate look like they should be a
list of key-value pairs. However, they're actually a list of lists of
key-value pairs. Previously we only looked at the first element of each
sublist and the vast majority of certificates only have one element per
sublist.
However, it's possible to have multiple elements and some 360
certificates from the “Pilot” log are so constructed.
This change causes all elements of the sublists to be processed.