Rename StoreConst to ValAndOff so we can use it for other ops.
Make ValAndOff print nicely.
Add some notes & checks related to my aborted attempt to
implement combined CMP+load ops.
Change-Id: I2f901d12d42bc5a82879af0334806aa184a97e27
Reviewed-on: https://go-review.googlesource.com/18947
Run-TryBot: David Chase <drchase@google.com>
Reviewed-by: David Chase <drchase@google.com>
The old write barriers used _nostore versions, which
don't work for Ian's cgo checker. Instead, we adopt the
same write barrier pattern as the default compiler.
It's a bit trickier to code up but should be more efficient.
Change-Id: I6696c3656cf179e28f800b0e096b7259bd5f3bb7
Reviewed-on: https://go-review.googlesource.com/18941
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Compiling && and || expressions often leads to control
flow of the following form:
p:
If a goto b else c
b: <- p ...
x = phi(a, ...)
If x goto t else u
Note that if we take the edge p->b, then we are guaranteed
to take the edge b->t also. So in this situation, we might
as well go directly from p to t.
Change-Id: I6974f1e6367119a2ddf2014f9741fdb490edcc12
Reviewed-on: https://go-review.googlesource.com/18910
Reviewed-by: David Chase <drchase@google.com>
For each value that needs to be in a fixed register at the end of the
block, and try to pick that fixed register when the instruction
generating that value is scheduled (or restored from a spill).
Just used for end-of-block register requirements for now.
Fixed-register instruction requirements (e.g. shift in ecx) can be
added later. Also two-instruction constraints (input reg == output
reg) might be recorded in a similar manner.
Change-Id: I59916e2e7f73657bb4fc3e3b65389749d7a23fa8
Reviewed-on: https://go-review.googlesource.com/18774
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
The OpSB hack didn't quite work. We need to really
CSE these ops to make regalloc happy.
Change-Id: I9f4d7bfb0929407c84ee60c9e25ff0c0fbea84af
Reviewed-on: https://go-review.googlesource.com/19083
Run-TryBot: Keith Randall <khr@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Some tests make multiple Funcs per Config at once.
With value & block caching, we can't do that any more.
Change-Id: Ibdb60aa2fcf478f1726b3be0fcaa06b04433eb67
Reviewed-on: https://go-review.googlesource.com/19081
Reviewed-by: Keith Randall <khr@golang.org>
It is one of the slowest compiler phases right now, and we
run two of them.
Instead of using a map to make the initial partition, use a sort.
It is much less memory intensive.
Do a few optimizations to avoid work for size-1 equivalence classes.
Implement -N.
Change-Id: I1d2d85d3771abc918db4dd7cc30b0b2d854b15e1
Reviewed-on: https://go-review.googlesource.com/19024
Reviewed-by: David Chase <drchase@google.com>
The x86 backend automatically rewrites MOV $0, AX to
XOR AX, AX. That rewrite isn't ok when the flags register
is live across the MOV. Keep track of which moves care
about preserving flags, then disable this rewrite for them.
On x86, Prog.Mark was being used to hold the length of the
instruction. We already store that in Prog.Isize, so no
need to store it in Prog.Mark also. This frees up Prog.Mark
to hold a bitmask on x86 just like all the other architectures.
Update #12405
Change-Id: Ibad8a8f41fc6222bec1e4904221887d3cc3ca029
Reviewed-on: https://go-review.googlesource.com/18861
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Russ Cox <rsc@golang.org>
The conversion from -0.0 to +0.0 happens inside mpgetflt now.
The SSA code doesn't need this fix any more.
Change-Id: I6cd4f4a4e75b13cf284ebbb95b08af050ed9891c
Reviewed-on: https://go-review.googlesource.com/18942
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
Run-TryBot: Keith Randall <khr@golang.org>
Empty blocks are introduced to remove critical edges.
After regalloc, we can remove any of the added blocks
that are still empty.
Change-Id: I0b40e95ac3a6cc1e632a479443479532b6c5ccd9
Reviewed-on: https://go-review.googlesource.com/18833
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Break small structs up into their components so they
can be registerized.
Change StructSelect to use field indexes instead of
field offsets, as field offsets aren't unique in the
presence of zero-sized fields.
Change-Id: I2f1dc89f7fa58e1cf58aa1a32b238959d53f62e4
Reviewed-on: https://go-review.googlesource.com/18570
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Redo how we keep track of forward references when building SSA.
When the forward reference is resolved, update the Value node
in place.
Improve the phi elimination pass so it can simplify phis of phis.
Give SSA package access to decoded line numbers. Fix line numbers
for constant booleans.
Change-Id: I3dc9896148d260be2f3dd14cbe5db639ec9fa6b7
Reviewed-on: https://go-review.googlesource.com/18674
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Distinguish move/load/store ops. Unify some of this code a bit.
Reduces Mandelbrot slowdown with SSA from 58% to 12%.
Change-Id: I3276eaebcbcdd9de3f8299c79b5f25c0429194c4
Reviewed-on: https://go-review.googlesource.com/18677
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
If a failure occurs in SSA processing, we always report the
last line of the function we're compiling. Modify the callbacks
from SSA to the GC compiler so we can pass a line number back
and use it in Fatalf.
Change-Id: Ifbfad50d5e167e997e0a96f0775bcc369f5c397e
Reviewed-on: https://go-review.googlesource.com/18599
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
Fixes build on those systems.
Also fix printing of AVARLIVE.
Change-Id: I1b38cca0125689bc08e4e1bdd0d0c140b1ea079a
Reviewed-on: https://go-review.googlesource.com/18641
Reviewed-by: Russ Cox <rsc@golang.org>
Consider this code:
func f(*int)
func g() {
p := new(int)
f(p)
}
where f is an assembly function.
In general liveness analysis assumes that during the call to f, p is dead
in this frame. If f has retained p, p will be found alive in f's frame and keep
the new(int) from being garbage collected. This is all correct and works.
We use the Go func declaration for f to give the assembly function
liveness information (the arguments are assumed live for the entire call).
Now consider this code:
func h1() {
p := new(int)
syscall.Syscall(1, 2, 3, uintptr(unsafe.Pointer(p)))
}
Here syscall.Syscall is taking the place of f, but because its arguments
are uintptr, the liveness analysis and the garbage collector ignore them.
Since p is no longer live in h once the call starts, if the garbage collector
scans the stack while the system call is blocked, it will find no reference
to the new(int) and reclaim it. If the kernel is going to write to *p once
the call finishes, reclaiming the memory is a mistake.
We can't change the arguments or the liveness information for
syscall.Syscall itself, both for compatibility and because sometimes the
arguments really are integers, and the garbage collector will get quite upset
if it finds an integer where it expects a pointer. The problem is that
these arguments are fundamentally untyped.
The solution we have taken in the syscall package's wrappers in past
releases is to insert a call to a dummy function named "use", to make
it look like the argument is live during the call to syscall.Syscall:
func h2() {
p := new(int)
syscall.Syscall(1, 2, 3, uintptr(unsafe.Pointer(p)))
use(unsafe.Pointer(p))
}
Keeping p alive during the call means that if the garbage collector
scans the stack during the system call now, it will find the reference to p.
Unfortunately, this approach is not available to users outside syscall,
because 'use' is unexported, and people also have to realize they need
to use it and do so. There is much existing code using syscall.Syscall
without a 'use'-like function. That code will fail very occasionally in
mysterious ways (see #13372).
This CL fixes all that existing code by making the compiler do the right
thing automatically, without any code modifications. That is, it takes h1
above, which is incorrect code today, and makes it correct code.
Specifically, if the compiler sees a foreign func definition (one
without a body) that has uintptr arguments, it marks those arguments
as "unsafe uintptrs". If it later sees the function being called
with uintptr(unsafe.Pointer(x)) as an argument, it arranges to mark x
as having escaped, and it makes sure to hold x in a live temporary
variable until the call returns, so that the garbage collector cannot
reclaim whatever heap memory x points to.
For now I am leaving the explicit calls to use in package syscall,
but they can be removed early in a future cycle (likely Go 1.7).
The rule has no effect on escape analysis, only on liveness analysis.
Fixes#13372.
Change-Id: I2addb83f70d08db08c64d394f9d06ff0a063c500
Reviewed-on: https://go-review.googlesource.com/18584
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Add new constant-flags opcodes. These can be generated from
comparisons that we know the result of, like x&31 < 32.
Constant-fold the constant-flags opcodes into all flag users.
Reorder some CMPxconst args so they read in the comparison direction.
Reorg deadcode removal a bit - it needs to remove the OpCopy ops it
generates when strength-reducing Phi ops. So it needs to splice out all
the dead blocks and do a copy elimination before it computes live
values.
Change-Id: Ie922602033592ad8212efe4345394973d3b94d9f
Reviewed-on: https://go-review.googlesource.com/18267
Run-TryBot: Keith Randall <khr@golang.org>
Reviewed-by: David Chase <drchase@google.com>
In code that does:
var x, z int32
var y int64
z = phi(x, int32(y))
We silently drop the int32 cast because truncation is a no-op.
The phi operation needs to make sure it uses the size of the
phi, not the size of its arguments, when generating spills.
Change-Id: I1f7baf44f019256977a46fdd3dad1972be209042
Reviewed-on: https://go-review.googlesource.com/18390
Reviewed-by: David Chase <drchase@google.com>
The fucomi* opcodes were only introduced for the Pentium Pro.
They do not exist for an MMX Pentium. Use the fucom* instructions
instead and move the condition codes from the fp flags register to
the integer flags register explicitly.
The use of fucomi* opcodes in ggen.go was introduced in 1.5 (CL 8738).
The bad ops were generated for 64-bit floating-point comparisons.
The use of fucomi* opcodes in gsubr.go dates back to at least 1.1.
The bad ops were generated for float{32,64} to uint64 conversions.
Fixes#13923
Change-Id: I5290599f5edea8abf8fb18036f44fa78bd1fc9e6
Reviewed-on: https://go-review.googlesource.com/18590
Reviewed-by: Minux Ma <minux@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Brief background on "why heap allocate". Things can be
forced to the heap for the following reasons:
1) address published, hence lifetime unknown.
2) size unknown/too large, cannot be stack allocated
3) multiplicity unknown/too large, cannot be stack allocated
4) reachable from heap (not necessarily published)
The bug here is a case of failing to enforce 4) when an
object Y was reachable from a heap allocation X forced
because of 3). It was found in the case of a closure
allocated within a loop (X) and assigned to a variable
outside the loop (multiplicity unknown) where the closure
also captured a map (Y) declared outside the loop (reachable
from heap). Note the variable declared outside the loop (Y)
is not published, has known size, and known multiplicity
(one). The only reason for heap allocation is that it was
reached from a heap allocated item (X), but because that was
not forced by publication, it has to be tracked by loop
level, but escape-loop level was not tracked and thus a bug
results.
The fix is that when a heap allocation is newly discovered,
use its looplevel as the minimum loop level for downstream
escape flooding.
Every attempt to generalize this bug to X-in-loop-
references-Y-outside loop succeeded, so the fix was aimed
to be general. Anywhere that loop level forces heap
allocation, the loop level is tracked. This is not yet
tested for all possible X and Y, but it is correctness-
conservative and because it caused only one trivial
regression in the escape tests, it is probably also
performance-conservative.
The new test checks the following:
1) in the map case, that if fn escapes, so does the map.
2) in the map case, if fn does not escape, neither does the map.
3) in the &x case, that if fn escapes, so does &x.
4) in the &x case, if fn does not escape, neither does &x.
Fixes#13799.
Change-Id: Ie280bef2bb86ec869c7c206789d0b68f080c3fdb
Reviewed-on: https://go-review.googlesource.com/18234
Run-TryBot: David Chase <drchase@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
There are fewer special cases this way: the import map applies
to all import paths, not just the ones not spelled "unsafe".
This is also consistent with what the code in cmd/go and go/build expects.
They make no exception for "unsafe".
For #13703.
Change-Id: I622295261ca35a6c1e83e8508d363bddbddb6c0a
Reviewed-on: https://go-review.googlesource.com/18438
Reviewed-by: Ian Lance Taylor <iant@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
Reviewed-by: Robert Griesemer <gri@golang.org>
Adding the evconst(n) call for OANDAND and OOROR in
golang.org/cl/18262 was originally just to parallel the above iscmp
branch, but upon further inspection it seemed odd that removing it
caused test/fixedbugs/issue6671.go's
var b mybool
// ...
b = bool(true) && true // ERROR "cannot use"
to start failing (i.e., by not emitting the expected "cannot use"
error).
The problem is that evconst(n)'s settrue and setfalse paths always
reset n.Type to idealbool, even for logical operators where n.Type
should preserve the operand type. Adding the evconst(n) call for
OANDAND/OOROR inadvertantly worked around this by turning the later
evconst(n) call at line 2167 into a noop, so the "n.Type = t"
assignment at line 739 would preserve the operand type.
However, that means evconst(n) was still clobbering n.Type for ONOT,
so declarations like:
const _ bool = !mybool(true)
were erroneously accepted.
Update #13821.
Change-Id: I18e37287f05398fdaeecc0f0d23984e244f025da
Reviewed-on: https://go-review.googlesource.com/18362
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Russ Cox <rsc@golang.org>
Forgot to reset these masks before each merge edge is processed.
Change-Id: I2f593189b63f50a1cd12b2dd4645ca7b9614f1f3
Reviewed-on: https://go-review.googlesource.com/18223
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: David Chase <drchase@google.com>
Added a format option to inhibit output of .Note field in
printing, and enabled that option during export.
Added test.
Fixes#13777.
Change-Id: I739f9785eb040f2fecbeb96d5a9ceb8c1ca0f772
Reviewed-on: https://go-review.googlesource.com/18217
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: David Chase <drchase@google.com>
Reorder how register & stack allocation is done. We used to allocate
registers, then fix up merge edges, then allocate stack slots. This
lead to lots of unnecessary copies on merge edges:
v2 = LoadReg v1
v3 = StoreReg v2
If v1 and v3 are allocated to the same stack slot, then this code is
unnecessary. But at regalloc time we didn't know the homes of v1 and
v3.
To fix this problem, allocate all the stack slots before fixing up the
merge edges. That way, we know what stack slots values use so we know
what copies are required.
Use a good technique for shuffling values around on merge edges.
Improves performance of the go1 TimeParse benchmark by ~12%
Change-Id: I731f43e4ff1a7e0dc4cd4aa428fcdb97812b86fa
Reviewed-on: https://go-review.googlesource.com/17915
Reviewed-by: David Chase <drchase@google.com>
We used to compile everything with SSA and then decide whether
to use the result or not. It was useful when we were working
on coverage without much regard for correctness, but not so much now.
Instead, let's decide what we're going to compile and go through
the SSA compiler for only those functions.
TODO: next CL: get rid of all the UnimplementedF stuff.
Change-Id: If629addd8b62cd38ef553fd5d835114137885ce0
Reviewed-on: https://go-review.googlesource.com/17763
Reviewed-by: David Chase <drchase@google.com>
Fixes#12411.
Change-Id: I2202a754c7750e3b2119e3744362c98ca0d2433e
Reviewed-on: https://go-review.googlesource.com/17818
Reviewed-by: Keith Randall <khr@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
This simply copies the current version of math/big into the
compiler directory. The change was created automatically by
running cmd/compile/internal/big/vendor.bash. No other manual
changes.
Change-Id: Ica225d196b3ac10dfd9d4dc1e4e4ef0b22812ff9
Reviewed-on: https://go-review.googlesource.com/17900
Run-TryBot: Robert Griesemer <gri@golang.org>
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
After fixing #13587, I noticed that the "OAS2FUNC in disguise" block
looked like it probably needed write barriers too. However, testing
revealed the multi-value "return f()" case was already being handled
correctly.
It turns out this block is dead code due to "return f()" already being
transformed into "t1, t2, ..., tN := f(); return t1, t2, ..., tN" by
orderstmt when f is a multi-valued function.
Updates #13587.
Change-Id: Icde46dccc55beda2ea5fd5fcafc9aae26cec1552
Reviewed-on: https://go-review.googlesource.com/17759
Run-TryBot: Matthew Dempsky <mdempsky@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Reviewed-by: Keith Randall <khr@golang.org>
With the separate flagalloc pass, it should be fine to
allow CSE of control values. The worst that can happen
is that the comparison gets un-CSEd by flagalloc.
Fix bug in flagalloc where flag restores were getting
clobbered by rematerialization during register allocation.
Change-Id: If476cf98b69973e8f1a8eb29441136dd12fab8ad
Reviewed-on: https://go-review.googlesource.com/17760
Reviewed-by: David Chase <drchase@google.com>
Run-TryBot: Keith Randall <khr@golang.org>
Spilling/restoring flag values is a pain to do during regalloc.
Instead, allocate the flag register in a separate pass. Regalloc then
operates normally on any flag recomputation instructions.
Change-Id: Ia1c3d9e6eff678861193093c0b48a00f90e4156b
Reviewed-on: https://go-review.googlesource.com/17694
Reviewed-by: David Chase <drchase@google.com>
In particular, we can initialize globals with them at link time instead
of generating code for them in an init() function. Less code, less
startup cost.
But the real reason for this change is binary size. This change reduces
the binary size of hello world by ~4%.
The culprit is fmt.ssFree, a global variable which is a sync.Pool of
scratch scan states. It is initalized with a captureless closure as the
pool's New action. That action in turn references all the scanf code.
If you never call any of the fmt.Scanf* routines, ssFree is never used.
But before this change, ssFree is still referenced by fmt's init
function. That keeps ssFree and all the code it references in the
binary. With this change, ssFree is initialized at link time. As a
result, fmt.init never mentions ssFree. If you don't call fmt.Scanf*,
ssFree is unreferenced and it and the scanf code are not included.
This change is an easy fix for what is generally a much harder problem,
the unnecessary initializing of unused globals (and retention of code
that they reference). Ideally we should have separate init code for
each global and only include that code if the corresponding global is
live. (We'd need to make sure that the initializing code has no side
effects, except on the global being initialized.) That is a much harder
change.
Update #6853
Change-Id: I19d1e33992287882c83efea6ce113b7cfc504b67
Reviewed-on: https://go-review.googlesource.com/17398
Reviewed-by: Russ Cox <rsc@golang.org>
Run-TryBot: Russ Cox <rsc@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
This code used to be necessary because of the error messages generated
by the YACC-based parser, but they're no longer relevant under the new
recursive descent parser:
- LBRACE no longer exists, so "{ or {" can never occur.
- The parser never generates error messages about "@" or "?" now
(except in import sections, where they're actually legitimate).
- The s/LLITERAL/litbuf/ substitution is handled in p.syntax_error.
Change-Id: Id39f747e4aa492c5830d14a47b161920bd4589ad
Reviewed-on: https://go-review.googlesource.com/17690
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
When using GOEXPERIMENT=fieldtrack, we can see AUSEFIELD instructions.
We generally want to ignore them.
No tests because as far as I can tell there are no tests for
GOEXPERIMENT=fieldtrack.
Change-Id: Iee26f25592158e5db691a36cf8d77fc54d051314
Reviewed-on: https://go-review.googlesource.com/17610
Reviewed-by: David Symonds <dsymonds@golang.org>
Run-TryBot: Ian Lance Taylor <iant@golang.org>
TryBot-Result: Gobot Gobot <gobot@golang.org>
Another (historic) artifact due to partially resolving symbols too early.
Fixes#13539.
Change-Id: Ie720c491cfa399599454f384b3a9735e75d4e8f1
Reviewed-on: https://go-review.googlesource.com/17600
Run-TryBot: Robert Griesemer <gri@golang.org>
Reviewed-by: Damien Neil <dneil@google.com>
TryBot-Result: Gobot Gobot <gobot@golang.org>
