PR go/65349
runtime: Don't call malloc from __go_file_line callback.
When crashing, we call runtime_printcreatedby which calls
__go_file_line which used to call the Go malloc. If we are
crashing due to a signal due to heap corruption of some sort,
the GO malloc lock might already be held, leading to a crash
within a crash. Avoid that by assuming that the libbacktrace
strings will stick around, as we already do in go-callers.c.
From-SVN: r221291
From Rainer Orth.
The recent godump changes broke Solaris 11.1+ bootstrap in
libgo: before, gen-sysinfo.so had
type _zone_net_addr_t struct { zna_family uint16; zna_plen uint16; zna_addru struct { znau_addr6 _in6_addr; }; }
which was filtered out by mksysinfo.sh due to the use of
_in6_addr.
After the change, there's now
type _zone_net_addr_t struct { zna_family uint16; zna_plen uint16; zna_addru struct { znau_addr6 [16]byte; Godump_0_align [0]uint32; }; }
instead, not filtered, but added a second time by the
_zone_net_addr_t code in mksysinfo.sh, which leads to
redefinition warnings/errors.
Simply removing the old _zone_net_addr_t fragment fixes this
and restores bootstrap.
From-SVN: r220481
Add memprofilerate as a value recognized
in the GODEBUG env var. The value provided
is used as the new setting for
runtime.MemProfileRate, allowing the user
to adjust memory profiling.
From-SVN: r220470
PR go/64683
runtime/pprof: Let memory profiler test pass if value not collected.
Since gccgo's GC is not precise, the transient value may not
be collected. Let the regexp match that case as well.
From-SVN: r219900
Change from using __go_set_closure to passing the closure
value in the static chain field. Uses new backend support for
setting the closure chain in a call from C via
__builtin_call_with_static_chain. Uses new support in libffi
for Go closures.
The old architecture specific support for reflect.MakeFunc is
removed, replaced by the libffi support.
All work done by Richard Henderson.
* go-gcc.cc (Gcc_backend::call_expression): Add chain_expr argument.
(Gcc_backend::static_chain_variable): New method.
From-SVN: r219776
The version from the master repository seems to be appropriate
for OpenSolaris but not for Solaris itself. Solaris 11.2
proper does not define TCP_KEEPIDLE or TCP_KEEPINTVL.
From-SVN: r219749
A recent libffi upgrade caused the reflect test to fail on
386. The problem case is a function that returns an empty
struct--a struct with no fields. The libffi library does not
recognize the existence of empty structs, presumably since
they can't happen in C. To work around this, the Go interface
to the libffi library changes an empty struct to void. This
normally works fine, but with the new libffi upgrade it fails
for a function that returns an empty struct. On 386 a
function that returns a struct is expected to pop the hidden
pointer when it returns. So when we convert an empty struct
to void, libffi is calling a function that pops the hidden
pointer but does not expect that to happen.
In the older version of libffi, this didn't matter, because
the libffi code for 386 used a frame pointer, so the fact that
the stack pointer was wonky when the function returned was
ignored as the stack pointer was immediately replaced by the
saved frame pointer. In the newer version of libffi, the 386
code is more efficient and does not use a frame pointer, and
therefore it matters whether libffi expects the function to
pop the hidden pointer or not.
This patch changes libgo to convert an empty to a struct with
a single field of type void. This seems to be enough to get
the test cases working again.
Of course the real fix would be to change libffi to handle
empty types, but as libffi uses size == 0 as a marker for an
uninitialized type, that would be a non-trivial change.
From-SVN: r219701
This upgrades all of libgo other than the runtime package to
the Go 1.4 release. In Go 1.4 much of the runtime was
rewritten into Go. Merging that code will take more time and
will not change the API, so I'm putting it off for now.
There are a few runtime changes anyhow, to accomodate other
packages that rely on minor modifications to the runtime
support.
The compiler changes slightly to add a one-bit flag to each
type descriptor kind that is stored directly in an interface,
which for gccgo is currently only pointer types. Another
one-bit flag (gcprog) is reserved because it is used by the gc
compiler, but gccgo does not currently use it.
There is another error check in the compiler since I ran
across it during testing.
gotools/:
* Makefile.am (go_cmd_go_files): Sort entries. Add generate.go.
* Makefile.in: Rebuild.
From-SVN: r219627
Don't look it up on PATH. This lets a go tool built as part
of a GCC build use the gccgo from the same build.
Also pass -c when assembling a .s file with gccgo.
From-SVN: r219358
Fix an unusual C to Go callback case. Newly created C threads
call into Go code, forcing the Go code to allocate new M and G
structures. While executing Go code, the stack is split. The
Go code then returns. Returning from a Go callback is treated
as entering a system call, so the G gcstack field is set to
point to the Go stack. In this case, though, we were called
from a newly created C thread, so we drop the extra M and G
structures. The C thread then exits.
Then a new C thread calls into Go code, reusing the previously
created M and G. The Go code requires a larger stack frame,
causing the old stack segment to be unmapped and a new stack
segment allocated. At this point the gcstack field is
pointing to the old stack segment.
Then a garbage collection occurs. The garbage collector sees
that the gcstack field is not nil, so it scans it as the first
stack segment. Unfortunately it points to memory that was
unmapped. So the program crashes.
The fix is simple: when handling extra G structures created
for callbacks from new C threads, clear the gcstack field.
From-SVN: r218699
Different compilers may have different release cadences or ABI
incompatibilities, so it does not make sense to use the same library
name for runtime libraries intended for different compilers.
This CL causes a libgo built by llgo to receive the name libgo-llgo.
Likewise, libgobegin is named libgobegin-llgo.
From-SVN: r217583
From Dominik Vogt.
* libgo/go/syscall/libcall_linux_s390.go: New file for s390 support.
* libgo/go/syscall/syscall_linux_s390.go: Ditto.
* libgo/go/syscall/libcall_linux_s390x.go: New file for s390x support.
* libgo/go/syscall/syscall_linux_s390x.go: Ditto.
* libgo/go/runtime/pprof/pprof.go (printStackRecord): Support s390 and
s390x.
* libgo/runtime/runtime.c (runtime_cputicks): Add support for s390 and
s390x
* libgo/mksysinfo.sh: Ditto.
(upcase_fields): New helper function
* libgo/go/debug/elf/file.go (applyRelocations): Implement relocations
on s390x.
(applyRelocationsS390x): Ditto.
(DWARF): Ditto.
* libgo/go/debug/elf/elf.go (R_390): New constants for S390 relocations.
(r390Strings): Ditto.
(String): Helper function for S390 relocations.
(GoString): Ditto.
* libgo/go/reflect/makefuncgo_s390.go: New file.
(S390MakeFuncStubGo): Implementation of s390 abi.
* libgo/go/reflect/makefuncgo_s390x.go: New file.
(S390xMakeFuncStubGo): Implementation of s390x abi.
* libgo/go/reflect/makefunc_s390.c: New file.
(makeFuncStub): s390 and s390x specific implementation of function.
* libgo/go/reflect/makefunc.go
(MakeFunc): Add support for s390 and s390x.
(makeMethodValue): Ditto.
(makeValueMethod): Ditto.
* libgo/Makefile.am (go_reflect_makefunc_s_file): Ditto.
(go_reflect_makefunc_file): Ditto.
* libgo/go/reflect/makefunc_dummy.c: Ditto.
* libgo/runtime/runtime.h (__go_makefunc_can_recover): Export prototype
for use in makefunc_s390.c.
(__go_makefunc_returning): Ditto.
* libgo/go/syscall/exec_linux.go (forkAndExecInChild): Fix order of the
arguments of the clone system call for s390[x].
* libgo/configure.ac (is_s390): New variable.
(is_s390x): Ditto
(LIBGO_IS_S390): Ditto.
(LIBGO_IS_S390X): Ditto.
(GOARCH): Support s390 and s390x.
* libgo/go/go/build/build.go (cgoEnabled): Ditto.
* libgo/go/go/build/syslist.go (goarchList): Ditto.
From-SVN: r217106
Backport https://codereview.appspot.com/155450044 from the
master Go library. Original description:
I came across this while debugging a GC problem in gccgo.
There is code in assignTo and cvtT2I that handles assignment
to all interface values. It allocates an empty interface even
if the real type is a non-empty interface. The fields are
then set for a non-empty interface, but the memory is recorded
as holding an empty interface. This means that the GC has
incorrect information.
This is extremely unlikely to fail, because the code in the GC
that handles empty interfaces looks like this:
obj = nil;
typ = eface->type;
if(typ != nil) {
if(!(typ->kind&KindDirectIface) || !(typ->kind&KindNoPointers))
obj = eface->data;
In the current runtime the condition is always true--if
KindDirectIface is set, then KindNoPointers is clear--and we
always want to set obj = eface->data. So the question is what
happens when we incorrectly store a non-empty interface value
in memory marked as an empty interface. In that case
eface->type will not be a *rtype as we expect, but will
instead be a pointer to an Itab. We are going to use this
pointer to look at a *rtype kind field. The *rtype struct
starts out like this:
type rtype struct {
size uintptr
hash uint32 // hash of type; avoids computation in hash tables
_ uint8 // unused/padding
align uint8 // alignment of variable with this type
fieldAlign uint8 // alignment of struct field with this type
kind uint8 // enumeration for C
An Itab always has at least two pointers, so on a
little-endian 64-bit system the kind field will be the high
byte of the second pointer. This will normally be zero, so
the test of typ->kind will succeed, which is what we want.
On a 32-bit system it might be possible to construct a failing
case by somehow getting the Itab for an interface with one
method to be immediately followed by a word that is all ones.
The effect would be that the test would sometimes fail and the
GC would not mark obj, leading to an invalid dangling
pointer. I have not tried to construct this test.
I noticed this in gccgo, where this error is much more likely
to cause trouble for a rather random reason: gccgo uses a
different layout of rtype, and in gccgo the kind field happens
to be the low byte of a pointer, not the high byte.
From-SVN: r216489
We want to create goroutines with a small stack, at least on
systems where split stacks are supported. We don't need to
create threads with a small stack.
From-SVN: r216353
PR go/60406
runtime: Check callers in can_recover if return address doesn't match.
Also use __builtin_extract_return_address and tighten up the
checks in FFI code.
Fixes PR 60406.
From-SVN: r216003
PR go/61877
refect: fix direct call of variadic method value
As reported in bug https://gcc.gnu.org/bugzilla/show_bug.cgi?id=61877
gcc mainline has regressed in this. This CL adds the tests proposed
for the main Go repository:
https://codereview.appspot.com/151280043/https://codereview.appspot.com/152060043/
restores the code from the amd64/386 path that makes this work and
was lost when the Go 1.3 stdlib was merged and changes the FFI path
to call into the same helper code as the amd64/386 path.
I've only tested this on amd64 but I did test a version that was
patched to unconditionally take the FFI path.
From-SVN: r215859
This supports clang, which by default issues warnings about
unused command line arguments, a habit that interacts poorly
with configure scripts.
From-SVN: r215699
If the compiler inlines this function into kickoff, it may reuse
the TLS block address to load g. However, this is not necessarily
correct, as the call to g->entry in kickoff may cause the TLS
address to change. If the wrong value is loaded for g->status in
runtime_goexit, it may cause a runtime panic.
By marking the function as noinline we prevent the compiler from
reusing the TLS address.
From-SVN: r215484
The Go frontend passes closures through to functions using the
functions __go_set_closure and __go_get_closure. The
expectation is that there are no function calls between
set_closure and get_closure. However, it turns out that there
can be function calls if some of the function arguments
require type conversion to an interface type. Converting to
an interface type can allocate memory, and that can in turn
trigger a garbage collection, and that can in turn call pool
cleanup functions that may call __go_set_closure. So the
called function can see the wrong closure value, which is bad.
This patch fixes the problem in two different ways. First, we
move all type conversions in function arguments into temporary
variables so that they can not appear before the call to
__go_set_closure. (This required shifting the flatten phase
after the simplify_thunk phase, since the latter expects to
work with unconverted argument types.) Second, we fix the
memory allocation function to preserve the closure value
across any possible garbage collection.
A test case is the libgo database/sql check run with the
environment variable GOGC set to 1.
From-SVN: r213932
PR other/61895
runtime: Ignore small argv[0] file for backtrace.
Reportedly in some cases Docker starts processes with argv[0]
pointing to an empty file. That would cause libgo to pass
that empty file to libbacktrace, which would then fail to do
any backtraces. Everything should work fine if libbacktrace
falls back to /proc/self/exe.
This patch to libgo works around the problem by ignoring
argv[0] if it is a small file, or if stat fails. This is not
a perfect fix but it's an unusual problem.
From-SVN: r213513
This variable is unused apparently as a result of local changes.
gccgo accepts this variable declaration, but other frontends may not.
From-SVN: r212873
PR go/61620
runtime: Don't free tiny blocks in map deletion.
The memory allocator now has a special case for tiny blocks
(smaller than 16 bytes) and they can not be explicitly freed.
From-SVN: r212233
This introduces the "bench" build target, which can be used to run
all benchmarks.
It is also possible to run subsets of benchmarks with the
"package/check" build targets by setting GOBENCH to a matching regex.
From-SVN: r212212
PR go/52583
runtime: Stop backtrace at a few recognized functions.
On x86_64 Solaris the makecontext function does not properly
indicate that it is at the top of the stack. Attempting to
unwind the stack past a call to makecontext tends to crash.
This patch changes libgo to look for certain functions that
are always found at the top of the stack, and to stop
unwinding when it reaches one of those functions. There is
never anything interesting past these functions--that is,
there is never any code written by the user.
From-SVN: r211640
PR go/61498
runtime: Always set gcnext_sp to pointer-aligned address.
The gcnext_sp field is only used on systems that do not use
split stacks. It marks the bottom of the stack for the
garbage collector. This change makes sure that the stack
bottom is always aligned to a pointer value.
Previously the garbage collector would align all the addresses
that it scanned, but it now expects them to be aligned before
scanning.
From-SVN: r211639
This revision was committed January 7, 2014. The next
revision deleted runtime/mfinal.c. That will be done in a
subsequent merge.
This merge changes type descriptors to add a zero field,
pointing to a zero value for that type. This is implemented
as a common variable.
* go-gcc.cc (Gcc_backend::implicit_variable): Add is_common and
alignment parameters. Permit init parameter to be NULL.
From-SVN: r211249
This adds the --without-libatomic configure option, which is useful for building libgo
with a non-gcc compiler.
It disables libgo's dependency on libatomic. This
is useful for platforms where it is known that the libatomic runtime
functions are not required, or where the compiler automatically
provides an implementation of them.
From-SVN: r211065
LLVM's code generator does not currently support split stacks for vararg
functions, so we disable split stacks for the only function that uses this
feature under Clang. This appears to be OK as long as:
- this function only calls non-inlined, internal-linkage (hence no dynamic
loader) functions compiled with split stacks (i.e. go_vprintf), which can
allocate more stack space as required;
- this function itself does not occupy more than BACKOFF bytes of stack space
(see libgcc/config/i386/morestack.S).
These conditions are currently known to be satisfied by Clang on x86-32 and
x86-64. Note that signal handlers receive slightly less stack space than they
would normally do if they happen to be called while this function is being
run. If this turns out to be a problem we could consider increasing BACKOFF.
From-SVN: r211037
This includes the use of __complex and __builtin_ functions where
unprefixed entities would suffice, and the use of a union for
bit-casting between types.
From-SVN: r211036
PR go/60931
runtime: Fix garbage collector issue with non 4kB system page size
The go garbage collector tracks memory in terms of 4kB pages.
Most of the code checks getpagesize() at runtime and does the
right thing.
On a 64kB ppc64 box I see SEGVs in long running processes
which has been diagnosed as a bug in scavengelist.
scavengelist does a madvise(MADV_DONTNEED) without rounding
the arguments to the system page size. A strace of one of the
failures shows the problem:
madvise(0xc211030000, 4096, MADV_DONTNEED) = 0
The kernel rounds the length up to 64kB and we mark 60kB of
valid data as no longer needed.
Round start up to a system page and end down before calling
madvise.
From-SVN: r209777
A gccgo language extension allows a function to be declared multiple
times. Avoid the use of this extension by dedeplicating declarations
in mksyscall.awk.
From-SVN: r209508
Avoid the use of a gccgo language extension which allows unsafe.Sizeof
to accept a type by passing an expression of the relevant type.
From-SVN: r209503
This patch fixes a rare but serious bug. The Go garbage
collector only examines Go stacks. When Go code calls a
function that is not written in Go, it first calls
syscall.Entersyscall. Entersyscall records the position of
the Go stack pointer and saves a copy of all the registers.
If the garbage collector runs while the thread is executing
the non-Go code, the garbage collector fetches the stack
pointer and registers from the saved location.
Entersyscall saves the registers using the getcontext
function. Unfortunately I didn't consider the possibility
that Entersyscall might itself change a register before
calling getcontext. This only matters for callee-saved
registers, as caller-saved registers would be visible on the
saved stack. And it only matters if Entersyscall is compiled
to save and modify a callee-saved register before it calls
getcontext. And it only matters if a garbage collection
occurs while the non-Go code is executing. And it only
matters if the only copy of a valid Go pointer happens to be
in the callee-saved register when Entersyscall is called.
When all those conditions are true, the Go pointer might get
collected incorrectly, leading to memory corruption.
This patch tries to avoid the problem by splitting
Entersyscall into two functions. The first is a simple
function that just calls getcontext and then calls the rest of
Entersyscall. This should fix the problem, provided the
simple Entersyscall function does not itself modify any
callee-saved registers before calling getcontext. That seems
to be true on the systems I checked. But since the argument
to getcontext is an offset from a TLS variable, it won't be
true on a system which needs to save callee-saved registers in
order to get the address of a TLS variable. I don't know why
any system would work that way, but I don't know how to rule
it out. I think that on any such system this will have to be
implemented in assembler. I can't put the ucontext_t
structure on the stack, because this function can not split
stacks, and the ucontext_t structure is large enough that it
could cause a stack overflow.
From-SVN: r208390
Before this, the heap location used on a 64-bit system was not
available to user-space on arm64, so the "32-bit" strategy ended up
being used. So use somewhere that is available, and for bonus points
is far away from where the kernel allocates address space by default.
From-SVN: r207977
The spans array is allocated in runtime_mallocinit. On a
32-bit system the number of entries in the spans array is
MaxArena32 / PageSize, which (2U << 30) / (1 << 12) == (1 << 19).
So we are allocating an array that can hold 19 bits for an
index that can hold 20 bits. According to the comment in the
function, this is intentional: we only allocate enough spans
(and bitmaps) for a 2G arena, because allocating more would
probably be wasteful.
But since the span index is simply the upper 20 bits of the
memory address, this scheme only works if memory addresses are
limited to the low 2G of memory. That would be OK if we were
careful to enforce it, but we're not. What we are careful to
enforce, in functions like runtime_MHeap_SysAlloc, is that we
always return addresses between the heap's arena_start and
arena_start + MaxArena32.
We generally get away with it because we start allocating just
after the program end, so we only run into trouble with
programs that allocate a lot of memory, enough to get past
address 0x80000000.
This changes the code that computes a span index to subtract
arena_start on 32-bit systems just as we currently do on
64-bit systems.
From-SVN: r206501
I am reliably informed that the architecture name and letter for the
plan9/inferno compilers for 64-bit ARM systems will be "arm64" and "7"
respectively, so let's get that bit in nice and early.
From Michael Hudson-Doyle.
https://codereview.appspot.com/34830045/
From-SVN: r206374
On Solaris, if you do a in-progress connect, and then the
server accepts and closes the socket, the client's later
attempt to complete the connect will fail with EINVAL. Handle
this case by assuming that the connect succeeded. This code
is weird enough that it is implemented as Solaris-only so that
it doesn't hide a real error on a different OS.
See http://golang.org/issue/6828.
From-SVN: r206232
PR go/59506
net: use DialTimeout in TestSelfConnect
Backported from master repository.
This avoids problems with systems that take a long time to
find out nothing is listening, while still testing for the
self-connect misfeature since a self-connect should be fast.
With this we may be able to remove the test for non-Linux
systems.
Tested (on GNU/Linux) by editing selfConnect in
tcpsock_posix.go to always return false and verifying that
TestSelfConnect then fails with and without this change.
Idea from Uros Bizjak.
From-SVN: r206224
When a 386 function returns a struct, it needs to return using
an rtd instruction that pops the hidden struct parameter off
the stack. That wasn't happening.
From-SVN: r205551
In particular this means that the names Getsockname returns are not
truncated to 26 characters.
Fixes issue 6829
https://codereview.appspot.com/31840043/
From-SVN: r205368
Fixes issue 6761
This simple change seems to work fine, slightly to my surprise.
This includes the tests I submitted to the main Go repository at
https://codereview.appspot.com/26570046
From-SVN: r205001
If cmd/go is rebuilt using -compiler gccgo the version of go/build that is linked into that cmd/go will not function properly as the list of file suffixes know as operating systems or architectures is incorrect.
From-SVN: r204794
The gccgo-specific iword function was checking v.kind, but for
a method value that is always Func. Fix to check v.typ.Kind()
instead.
From-SVN: r202670
This changes the compiler and runtime to not pass a closure
value as the last argument, but to instead pass it via
__go_set_closure and retrieve it via __go_get_closure. This
eliminates the need for function descriptor wrapper functions.
It will make it possible to retrieve the closure value in a
reflect.MakeFunc function.
From-SVN: r202233
A function that returns an interface type and returns a value
that requires memory allocation will try to allocate while
appearing to be in a syscall. This patch lets that work.
From-SVN: r201226
This fixes a problem on Solaris, where end is not defined in
the main program but comes from some shared library. This
only matters for 32-bit targets.
From-SVN: r201220
This changes the representation of a Go value of function type
from being a pointer to function code (like a C function
pointer) to being a pointer to a struct. The first field of
the struct points to the function code. The remaining fields,
if any, are the addresses of variables referenced in enclosing
functions. For each call to a function, the address of the
function descriptor is passed as the last argument.
This lets us avoid generating trampolines, and removes the use
of writable/executable sections of the heap.
From-SVN: r200181
PR go/56320
runtime: Support Solaris AMD64 in lfstack.
The address space layout is similar on SPARC64 and AMD64 when
running Solaris.
From-SVN: r196179
The mmap() call which reserves the arena should have MAP_NORESERVE
flag as in typical cases this memory will never be (fully) needed.
This matters in environments which do not do Linux style memory
overcommit, such as OpenIndiana/OpenSolaris/Solaris.
The MAP_NORESERVE flag does not exist on all operating systems
(for example FreeBSD). Therefore we define it to zero value in
case it does not exist.
Fixes issue 21.
From-SVN: r196088
PR go/56172
net: Fixes for select based pollster.
Make Close work properly, mainly for testing. Restart the
select if a descriptor is closed.
From-SVN: r195823
Uncovered by Uros Bizjak.
Before this patch the test would close the file descriptor but
not the os.File. When the os.File was GC'ed, the finalizer
would close the file descriptor again. That would cause
problems if the same file descriptor were returned by a later
call to open in another test.
On my system:
> GOGC=30 go test
--- FAIL: TestPassFD (0.04 seconds)
passfd_test.go:62: FileConn: dup: bad file descriptor
FAIL
From-SVN: r192854
Ian Lance Taylor
Richard Henderson
Richard Henderson
Ian Lance Taylor
Kyrylo Tkachov
Dominik Vogt
Chris Manghane
Uros Bizjak
Allan Sandfeld Jensen
Andreas Schwab