Use a lookup table to find the function which contains a pc. It is
faster than the old binary search. findfunc is used primarily for
stack copying and garbage collection.
benchmark old ns/op new ns/op delta
BenchmarkStackCopy 294746596 255400980 -13.35%
(findfunc is one of several tasks done by stack copy, the findfunc
time itself is about 2.5x faster.)
The lookup table is built at link time. The table grows the binary
size by about 0.5% of the text segment.
We impose a lower limit of 16 bytes on any function, which should not
have much of an impact. (The real constraint required is <=256
functions in every 4096 bytes, but 16 bytes/function is easier to
implement.)
Change-Id: Ic315b7a2c83e1f7203cd2a50e5d21a822e18fdca
Reviewed-on: https://go-review.googlesource.com/2097
Reviewed-by: Russ Cox <rsc@golang.org>
This implements support for calls to and from C in the ppc64 C ABI, as
well as supporting functionality such as an entry point from the
dynamic linker.
Change-Id: I68da6df50d5638cb1a3d3fef773fb412d7bf631a
Reviewed-on: https://go-review.googlesource.com/2009
Reviewed-by: Russ Cox <rsc@golang.org>
Cgo will need this for calls from C to Go and for handling signals
that may occur in C code.
Change-Id: I50cc4caf17cd142bff501e7180a1e27721463ada
Reviewed-on: https://go-review.googlesource.com/2008
Reviewed-by: Russ Cox <rsc@golang.org>
R13 is the C TLS pointer. Once we're calling to and from C code, if
we clobber R13 in our code, sigtramp won't know whether to get the
current g from REGG or from C TLS. The simplest solution is for Go
code to preserve the C TLS pointer. This is equivalent to what other
platforms do, except that on other platforms the TLS pointer is in a
special register.
Change-Id: I076e9cb83fd78843eb68cb07c748c4705c9a4c82
Reviewed-on: https://go-review.googlesource.com/2007
Reviewed-by: Minux Ma <minux@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
This implements the ELF relocations and dynamic linking tables
necessary to support internal linking on ppc64. It also marks ppc64le
ELF files as ABI v2; failing to do this doesn't seem to confuse the
loader, but it does confuse libbfd (and hence gdb, objdump, etc).
Change-Id: I559dddf89b39052e1b6288a4dd5e72693b5355e4
Reviewed-on: https://go-review.googlesource.com/2006
Reviewed-by: Russ Cox <rsc@golang.org>
Most ppc64 relocations come in six or more variants where the basic
relocation formula is the same, but which bits of the computed value
are installed where changes. Introduce the concept of "variants" for
internal relocations to support this. Since this applies to
architecture-independent relocation types like R_PCREL, we do this in
relocsym.
Currently there is only an identity variant. A later CL that adds
support for ppc64 ELF relocations will introduce more.
Change-Id: I0c5f0e7dbe5beece79cd24fe36267d37c52f1a0c
Reviewed-on: https://go-review.googlesource.com/2005
Reviewed-by: Russ Cox <rsc@golang.org>
ppc64 has a bunch of these.
Change-Id: I3b93ed2bae378322a8dec036b1681e520b56ff53
Reviewed-on: https://go-review.googlesource.com/2003
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Minux Ma <minux@golang.org>
ppc64 function symbols have both a global entry point and a local
entry point, where the difference is stashed in sym.other. We'll need
this information to generate calls to ELF ABI functions.
Change-Id: Ibe343923f56801de7ebec29946c79690a9ffde57
Reviewed-on: https://go-review.googlesource.com/2002
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Minux Ma <minux@golang.org>
Cache 2KB, 4KB, 8KB, and 16KB stacks. Larger stacks
will be allocated directly. There is no point in cacheing
32KB+ stacks as we ask for and return 32KB at a time
from the allocator.
Note that the minimum stack is 8K on windows/64bit and 4K on
windows/32bit and plan9. For these os/arch combinations,
the number of stack orders is less so that we have the same
maximum cached size.
Fixes#9045
Change-Id: Ia4195dd1858fb79fc0e6a91ae29c374d28839e44
Reviewed-on: https://go-review.googlesource.com/2098
Reviewed-by: Russ Cox <rsc@golang.org>
The ones at the end of M and G are just used to compute
their size for use in assembly. Generate the size explicitly.
The one at the end of itab is variable-sized, and at least one.
The ones at the end of interfacetype and uncommontype are not
needed, as the preceding slice references them (the slice was
originally added for use by reflect?).
The one at the end of stackmap is already accessed correctly,
and the runtime never allocates one.
Update #9401
Change-Id: Ia75e3aaee38425f038c506868a17105bd64c712f
Reviewed-on: https://go-review.googlesource.com/2420
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Fold in some startup randomness to make the hash vary across
different runs. This helps prevent attackers from choosing
keys that all map to the same bucket.
Also, reorganize the hash a bit. Move the *m1 multiply to after
the xor of the current hash and the message. For hash quality
it doesn't really matter, but for DDOS resistance it helps a lot
(any processing done to the message before it is merged with the
random seed is useless, as it is easily inverted by an attacker).
Update #9365
Change-Id: Ib19968168e1bbc541d1d28be2701bb83e53f1e24
Reviewed-on: https://go-review.googlesource.com/2344
Reviewed-by: Ian Lance Taylor <iant@golang.org>
The gc toolchain no longer includes a C compiler, so mentions of "6c"
can be removed or replaced by 6g as appropriate. Similarly, some cgo
functions that previously generated C source output no longer need to.
Change-Id: I1ae6b02630cff9eaadeae6f3176c0c7824e8fbe5
Reviewed-on: https://go-review.googlesource.com/2391
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Reader.Discard is the complement to Peek. It discards the next n bytes
of input.
We already have Reader.Buffered to see how many bytes of data are
sitting available in memory, and Reader.Peek to get that that buffer
directly. But once you're done with the Peek'd data, you can't get rid
of it, other than Reading it.
Both Read and io.CopyN(ioutil.Discard, bufReader, N) are relatively
slow. People instead resort to multiple blind ReadByte calls, just to
advance the internal b.r variable.
I've wanted this previously, several people have asked for it in the
past on golang-nuts/dev, and somebody just asked me for it again in a
private email. There are a few places in the standard library we'd use
it too.
Change-Id: I85dfad47704a58bd42f6867adbc9e4e1792bc3b0
Reviewed-on: https://go-review.googlesource.com/2260
Reviewed-by: Russ Cox <rsc@golang.org>
This CL only fixes the build, there are two failing tests:
RaceMapBigValAccess1 and RaceMapBigValAccess2
in runtime/race tests. I haven't investigated why yet.
Updates #9516.
Change-Id: If5bd2f0bee1ee45b1977990ab71e2917aada505f
Reviewed-on: https://go-review.googlesource.com/2401
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Use direct binary insertion instead of recursive calls to symMerge
when one of the blocks has only one element.
benchmark old ns/op new ns/op delta
BenchmarkStableString1K 421999 397629 -5.77%
BenchmarkStableInt1K 123422 120592 -2.29%
BenchmarkStableInt64K 9629094 9620200 -0.09%
BenchmarkStable1e2 123089 120209 -2.34%
BenchmarkStable1e4 39505228 36870029 -6.67%
BenchmarkStable1e6 8196612367 7630840157 -6.90%
Change-Id: I49905a909e8595cfa05920ccf9aa00a8f3036110
Reviewed-on: https://go-review.googlesource.com/2219
Reviewed-by: Robert Griesemer <gri@golang.org>
sysReserve doesn't actually reserve the full amount requested on
64-bit systems, because of problems with ulimit. Instead it checks
that it can get the first 64 kB and assumes it can grab the rest as
needed. This doesn't work well with the "let the kernel pick an address"
mode, so don't do that. Pick a high address instead.
Change-Id: I4de143a0e6fdeb467fa6ecf63dcd0c1c1618a31c
Reviewed-on: https://go-review.googlesource.com/2345
Reviewed-by: Rick Hudson <rlh@golang.org>
The line 'mp.schedlink = mnext' has an implicit write barrier call,
which needs a valid g. Move it above the setg(nil).
Change-Id: If3e86c948e856e10032ad89f038bf569659300e0
Reviewed-on: https://go-review.googlesource.com/2347
Reviewed-by: Minux Ma <minux@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
This test is doing pointer graph manipulation from C, and we
cannot support that with concurrent GC. The wbshadow mode
correctly diagnoses missing write barriers.
Disable the test in that mode for now. There is a bigger issue
behind it, namely SWIG, but for now we are focused on making
all.bash pass with wbshadow enabled.
Change-Id: I55891596d4c763e39b74082191d4a5fac7161642
Reviewed-on: https://go-review.googlesource.com/2346
Reviewed-by: Minux Ma <minux@golang.org>
Reviewed-by: Rick Hudson <rlh@golang.org>
There are two methods by which TLS clients signal the renegotiation
extension: either a special cipher suite value or a TLS extension.
It appears that I left debugging code in when I landed support for the
extension because there's a "+ 1" in the switch statement that shouldn't
be there.
The effect of this is very small, but it will break Firefox if
security.ssl.require_safe_negotiation is enabled in about:config.
(Although almost nobody does this.)
This change fixes the original bug and adds a test. Sadly the test is a
little complex because there's no OpenSSL s_client option that mirrors
that behaviour of require_safe_negotiation.
Change-Id: Ia6925c7d9bbc0713e7104228a57d2d61d537c07a
Reviewed-on: https://go-review.googlesource.com/1900
Reviewed-by: Russ Cox <rsc@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
SignPSS is documented as allowing opts to be nil, but actually
crashes in that case. This change fixes that.
Change-Id: Ic48ff5f698c010a336e2bf720e0f44be1aecafa0
Reviewed-on: https://go-review.googlesource.com/2330
Reviewed-by: Minux Ma <minux@golang.org>
First, call clearcheckmarks immediately after changing checkmark,
so that there is less time when the checkmark flag and the bitmap
are inconsistent. The tiny gap between the two lines is fine, because
the world is stopped. Before, the gap was much larger and included
such code as "go bgsweep()", which allocated.
Second, modify gcphase only when the world is stopped.
As written, gcscan_m was changing gcphase from 0 to GCscan
and back to 0 while other goroutines were running.
Another goroutine running at the same time might decide to
sleep, see GCscan, call gcphasework, and start "helping" by
scanning its stack. That's fine, except that if gcphase flips back
to 0 as the goroutine calls scanblock, it will start draining the
work buffers prematurely.
Both of these were found wbshadow=2 (and a lot of hard work).
Eventually that will run automatically, but right now it still
doesn't quite work for all.bash, due to mmap conflicts with
pthread-created threads.
Change-Id: I99aa8210cff9c6e7d0a1b62c75be32a23321897b
Reviewed-on: https://go-review.googlesource.com/2340
Reviewed-by: Rick Hudson <rlh@golang.org>
Found with GODEBUG=wbshadow=2 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: I5624b509a36650bce6834cf394b9da163abbf8c0
Reviewed-on: https://go-review.googlesource.com/2310
Reviewed-by: Rick Hudson <rlh@golang.org>
Use typedmemmove, typedslicecopy, and adjust reflect.call
to execute the necessary write barriers.
Found with GODEBUG=wbshadow=2 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: Iec5b5b0c1be5589295e28e5228e37f1a92e07742
Reviewed-on: https://go-review.googlesource.com/2312
Reviewed-by: Keith Randall <khr@golang.org>
These depend on storing arbitrary integer values using
pointer atomics, and we can't support that anymore.
Change-Id: I8cadd6d462c3eebdbe7078f43fe7c779fa8f52b3
Reviewed-on: https://go-review.googlesource.com/2311
Reviewed-by: Rick Hudson <rlh@golang.org>
A side effect of this change is that when assertI2T writes to the
memory for the T being extracted, it can use typedmemmove
for write barriers.
There are other ways we could have done this, but this one
finishes a TODO in package runtime.
Found with GODEBUG=wbshadow=2 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: Icbc8aabfd8a9b1f00be2e421af0e3b29fa54d01e
Reviewed-on: https://go-review.googlesource.com/2279
Reviewed-by: Keith Randall <khr@golang.org>
Found with GODEBUG=wbshadow=2 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: I1320d5340a9e421c779f24f3b170e33974e56e4f
Reviewed-on: https://go-review.googlesource.com/2278
Reviewed-by: Rick Hudson <rlh@golang.org>
Found with GODEBUG=wbshadow=2 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: Iea83d693480c2f3008b4e80d55821acff65970a6
Reviewed-on: https://go-review.googlesource.com/2277
Reviewed-by: Keith Randall <khr@golang.org>
Preparation for replacing many memmove calls in runtime
with typedmemmove, which is a clearer description of what
the routine is doing.
For the same reason, rename writebarriercopy to typedslicecopy.
Change-Id: I6f23bef2c2215509fefba175b16908f76dc7538c
Reviewed-on: https://go-review.googlesource.com/2276
Reviewed-by: Rick Hudson <rlh@golang.org>
Add write barrier to atomic operations manipulating pointers.
In general an atomic write of a pointer word may indicate racy accesses,
so there is no strictly safe way to attempt to keep the shadow copy
in sync with the real one. Instead, mark the shadow copy as not used.
Redirect sync/atomic pointer routines back to the runtime ones,
so that there is only one copy of the write barrier and shadow logic.
In time we might consider doing this for most of the sync/atomic
functions, but for now only the pointer routines need that treatment.
Found with GODEBUG=wbshadow=1 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: I852936b9a111a6cb9079cfaf6bd78b43016c0242
Reviewed-on: https://go-review.googlesource.com/2066
Reviewed-by: Rick Hudson <rlh@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
The Gobuf.g goroutine pointer is almost always updated by assembly code.
In one of the few places it is updated by Go code - func save - it must be
treated as a uintptr to avoid a write barrier being emitted at a bad time.
Instead of figuring out how to emit the write barriers missing in the
assembly manipulation, change the type of the field to uintptr, so that
it does not require write barriers at all.
Goroutine structs are published in the allg list and never freed.
That will keep the goroutine structs from being collected.
There is never a time that Gobuf.g's contain the only references
to a goroutine: the publishing of the goroutine in allg comes first.
Goroutine pointers are also kept in non-GC-visible places like TLS,
so I can't see them ever moving. If we did want to start moving data
in the GC, we'd need to allocate the goroutine structs from an
alternate arena. This CL doesn't make that problem any worse.
Found with GODEBUG=wbshadow=1 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: I85f91312ec3e0ef69ead0fff1a560b0cfb095e1a
Reviewed-on: https://go-review.googlesource.com/2065
Reviewed-by: Rick Hudson <rlh@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
Found with GODEBUG=wbshadow=1 mode.
Eventually that will run automatically, but right now
it still detects other missing write barriers.
Change-Id: Ic8624401d7c8225a935f719f96f2675c6f5c0d7c
Reviewed-on: https://go-review.googlesource.com/2064
Reviewed-by: Rick Hudson <rlh@golang.org>
Reviewed-by: Austin Clements <austin@google.com>
This is the detection code. It works well enough that I know of
a handful of missing write barriers. However, those are subtle
enough that I'll address them in separate followup CLs.
GODEBUG=wbshadow=1 checks for a write that bypassed the
write barrier at the next write barrier of the same word.
If a bug can be detected in this mode it is typically easy to
understand, since the crash says quite clearly what kind of
word has missed a write barrier.
GODEBUG=wbshadow=2 adds a check of the write barrier
shadow copy during garbage collection. Bugs detected at
garbage collection can be difficult to understand, because
there is no context for what the found word means.
Typically you have to reproduce the problem with allocfreetrace=1
in order to understand the type of the badly updated word.
Change-Id: If863837308e7c50d96b5bdc7d65af4969bf53a6e
Reviewed-on: https://go-review.googlesource.com/2061
Reviewed-by: Austin Clements <austin@google.com>
When constants were declared using unexported constants,
the type information was lost when those constants were filtered out.
This CL propagates the type information of unexported constants
so that it is available for display.
This is a follow-up to CL 144110044, which fixed this problem
specifically for _ constants.
Updates #5397.
Change-Id: I3f0c767a4007d88169a5634ab2870deea4e6a740
Reviewed-on: https://go-review.googlesource.com/2091
Reviewed-by: Robert Griesemer <gri@golang.org>
Noticed while investigating the speed of the runtime tests, as part
of debugging while Plan 9's runtime tests are timing out on GCE.
Change-Id: I95f5a3d967a0b45ec1ebf10067e193f51db84e26
Reviewed-on: https://go-review.googlesource.com/2283
Reviewed-by: Ian Lance Taylor <iant@golang.org>
The existing go code document did not link to the GOPATH documentation.
This will link to it, in hopes of making it more discoverable.
Change-Id: Ie4ded2fdce08f412e4acbcc93acdd76f5791b84a
Reviewed-on: https://go-review.googlesource.com/2265
Reviewed-by: Andrew Gerrand <adg@golang.org>
This reverts commit ab0535ae3f.
I think it will remain useful to distinguish code that must
run on a system stack from code that can run on either stack,
even if that distinction is no
longer based on the implementation language.
That is, I expect to add a //go:systemstack comment that,
in terms of the old implementation, tells the compiler,
to pretend this function was written in C.
Change-Id: I33d2ebb2f99ae12496484c6ec8ed07233d693275
Reviewed-on: https://go-review.googlesource.com/2275
Reviewed-by: Russ Cox <rsc@golang.org>
This CL splits the (ever growing) list of ca cert locations by major unix
platforms (darwin, windows and plan9 are already handled seperately).
Although it is clear the unix variants cannot manage to agree on some standard
locations, we can avoid to some extent an artificial ranking of priority
amongst the supported GOOSs.
* Split certFiles definition by GOOS
* Include NetBSD ca cert location
Fixes#9285
Change-Id: I6df2a3fddf3866e71033e01fce43c31e51b48a9e
Reviewed-on: https://go-review.googlesource.com/2208
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Reviewed-by: Andrew Gerrand <adg@golang.org>
This ensures that changing an image.YCbCr's Y values can't change its
chroma values, even after re-slicing up to capacity.
Change-Id: Icb626561522e336a3220e10f456c95330ae7db9e
Reviewed-on: https://go-review.googlesource.com/2209
Reviewed-by: Rob Pike <r@golang.org>
