Add timing/allocation information to each compiler pass for both the
console and html output.
Change-Id: I75833003b806a09b4fb1bbf63983258612cdb7b0
Reviewed-on: https://go-review.googlesource.com/14277
Reviewed-by: Keith Randall <khr@golang.org>
Store floats in AuxInt to reduce allocations.
Change-Id: I101e6322530b4a0b2ea3591593ad022c992e8df8
Reviewed-on: https://go-review.googlesource.com/14320
Reviewed-by: David Chase <drchase@google.com>
Reviewed-by: Keith Randall <khr@golang.org>
Store bools in AuxInt to reduce allocations.
Change-Id: Ibd26db67fca5e1e2803f53d7ef094897968b704b
Reviewed-on: https://go-review.googlesource.com/14276
Reviewed-by: Keith Randall <khr@golang.org>
Fallthrough return needs to be a return block before jumping
to the exit block.
Change-Id: I994de2064da5c326c9cade2c33cbb15bdbce5acb
Reviewed-on: https://go-review.googlesource.com/14256
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
It is confusing to have exceptional edges jump back into
real code. Distinguish return blocks, which execute acutal
code, and the exit block, which is a merge point for the regular
and exceptional return flow.
Prevent critical edge insertion from adding blocks on edges
into the exit block. These added blocks serve no purpose and
add a bunch of dead jumps to the assembly output. Furthermore,
live variable analysis is confused by these jumps.
Change-Id: Ifd69e6c00e90338ed147e7cb351b5100dc0364df
Reviewed-on: https://go-review.googlesource.com/14254
Reviewed-by: David Chase <drchase@google.com>
cse was incorrectly classifying -0.0 and 0.0 as equivalent. This lead
to invalid code as ssa uses PXOR -0.0, reg to negate a floating point.
Fixes math.
Change-Id: Id7eb10c71749eaed897f29b02c33891cf5820acf
Reviewed-on: https://go-review.googlesource.com/14205
Reviewed-by: Keith Randall <khr@golang.org>
complex128 was being treated as a complex64
Fixes math/cmplx.
Change-Id: I2996915b4cb6b94198d41cf08a30bd8531b9fec5
Reviewed-on: https://go-review.googlesource.com/14206
Reviewed-by: Keith Randall <khr@golang.org>
liblink was rewriting xor by a negative zero (used by SSA
for negation) as XORPS reg,reg.
Fixes strconv.
Change-Id: I627a0a7366618e6b07ba8f0ad0db0e102340c5e3
Reviewed-on: https://go-review.googlesource.com/14200
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
Reviewed-by: Keith Randall <khr@golang.org>
Modified to use the truncating conversion.
Fixes reflect.
Change-Id: I47bf3200abc2d2c662939a2a2351e2ff84168f4a
Reviewed-on: https://go-review.googlesource.com/14167
Reviewed-by: David Chase <drchase@google.com>
It does a XOR internally and clobbers flags.
Change-Id: Id6ef9219c4e6c3a2b5fc79c8d52bcfa30c148617
Reviewed-on: https://go-review.googlesource.com/14165
Reviewed-by: Keith Randall <khr@golang.org>
They do an AND or an OR internally, so they clobber flags.
Fixes#12441
Change-Id: I6c843bd268496bc13fc7e3c561d76619e961e8ad
Reviewed-on: https://go-review.googlesource.com/14180
Reviewed-by: Todd Neal <todd@tneal.org>
liblink rewrites MOV $0, reg into XOR reg, reg. Make MOVxconst clobber
flags so we don't generate invalid code in the unlikely case that it
matters. In testing, this change leads to no additional regenerated
flags due to a scheduling fix in CL14042.
Change-Id: I7bc1cfee94ef83beb2f97c31ec6a97e19872fb89
Reviewed-on: https://go-review.googlesource.com/14043
Reviewed-by: Keith Randall <khr@golang.org>
Still to do:
details, more testing corner cases. (e.g. negative zero)
Includes small cleanups for previous CL.
Note: complex division is currently done in the runtime,
so the division code here is apparently not yet necessary
and also not tested. Seems likely better to open code
division and expose the widening/narrowing to optimization.
Complex64 multiplication and division is done in wide
format to avoid cancellation errors; for division, this
also happens to be compatible with pre-SSA practice
(which uses a single complex128 division function).
It would-be-nice to widen for complex128 multiplication
intermediates as well, but that is trickier to implement
without a handy wider-precision format.
Change-Id: I595a4300f68868fb7641852a54674c6b2b78855e
Reviewed-on: https://go-review.googlesource.com/14028
Reviewed-by: Keith Randall <khr@golang.org>
Specifying types in rewrites for all subexpressions gets verbose
quickly. Allow opcodes to specify a default type which is used when
none is supplied explicitly.
Provide default types for a few easy opcodes. There are probably more
we can do, but this is a good start.
Change-Id: Iedc2a1a423cc3e2d4472640433982f9aa76a9f18
Reviewed-on: https://go-review.googlesource.com/14128
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
To reduce the number of spills, give any non-phi values whose argument
is the control the same priority as the control.
With mask.bash, this reduces regenerated flags from 603 to 240.
Change-Id: I26883d69e80357c56b343428fb528102b3f26e7a
Reviewed-on: https://go-review.googlesource.com/14042
Reviewed-by: Keith Randall <khr@golang.org>
Instead of trying to delete dead code as soon as we find it, just
mark it as dead using a PlainAndDead block kind. The deadcode pass
will do the real removal.
This way is somewhat more efficient because we don't need to mess
with successor and predecessor lists of all the dead blocks.
Fixes#12347
Change-Id: Ia42d6b5f9cdb3215a51737b3eb117c00bd439b13
Reviewed-on: https://go-review.googlesource.com/14033
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
This also fixes the encoding/gob TestTopLevelNilPointer failure.
Change-Id: I9b29a6fddffd51af305c685f3a8e2a0594bfeeab
Reviewed-on: https://go-review.googlesource.com/14032
Reviewed-by: Keith Randall <khr@golang.org>
addEdge had two identical implementations so make it an exported method
on Block.
Change-Id: I8c21655a9dc5074fefd7f63b2f5b51897571e608
Reviewed-on: https://go-review.googlesource.com/14040
Reviewed-by: Keith Randall <khr@golang.org>
Frontend has already rewriten fallthrough statements, we just need to
ignore them.
Change-Id: Iadf89b06a9f8f9e6e2e1e87c934f31add77a19a1
Reviewed-on: https://go-review.googlesource.com/14029
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
Fixes build. There may be no current block.
Change-Id: I0da8bab133dc207124556927698e7cd682e64ef5
Reviewed-on: https://go-review.googlesource.com/13989
Reviewed-by: Keith Randall <khr@golang.org>
Add a check to make sure value arguments dominate the value.
Phi elim output used to fail this test. When eliminating
redundant phis, phi elim was using one of the args and not
the ultimate source. For example:
b1: x = ...
-> b2 b3
b2: y = Copy x b3: z = Copy x
-> b4 -> b4
b4: w = phi y z
Phi elim eliminates w, but it used to replace w with (Copy y).
That's bad as b2 does not dominate b4. Instead we should
replace w with (Copy x).
Fixes#12347
Change-Id: I9f340cdabcda8e2e90359fb4f9250877b1fffe98
Reviewed-on: https://go-review.googlesource.com/13986
Reviewed-by: David Chase <drchase@google.com>
Still to do: arithmetic
Change-Id: I31fd23b34980c9ed4b4e304b8597134b2ba6ca5c
Reviewed-on: https://go-review.googlesource.com/14024
Reviewed-by: Keith Randall <khr@golang.org>
The old backend doesn't like ideal types,
and we want to reuse its stackmap generation.
OOROR and OANDAND expressions have ideal type.
The old backend didn't care,
because those expressions got rewritten away into
jumps before stackmap generation.
Fix the type during conversion.
Change-Id: I488e7499298d9aec71da39c202f6a7235935bc8d
Reviewed-on: https://go-review.googlesource.com/13980
Reviewed-by: Brad Fitzpatrick <bradfitz@golang.org>
The code previously always used AX causing errors. For now, just
switch off the type in order to at least generate valid code.
Change-Id: Iaf13120a24b62456b9b33c04ab31f2d5104b381b
Reviewed-on: https://go-review.googlesource.com/13943
Reviewed-by: David Chase <drchase@google.com>
Add blocks to remove critical edges, even when it looks like
there's no phi that requires it. Regalloc still likes to have
critical-edge-free graphs for other reasons.
Change-Id: I69f8eaecbc5d79ab9f2a257c2e289d60b18e43c8
Reviewed-on: https://go-review.googlesource.com/13933
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
The issue 12226 has been caused by the allocation of the same register
for the equality check of two byte values. The code in cgen.go freed the
register for the second operand before the allocation of the register
for the first operand.
Fixes#12226
Change-Id: Ie4dc33a488bd48a17f8ae9b497fd63c1ae390555
Reviewed-on: https://go-review.googlesource.com/13771
Reviewed-by: Russ Cox <rsc@golang.org>
Add a new function and generic operation to handle
bounds checking for slices. Unlike the index
bounds checking the index can be equal to the upper
bound.
Do gc-friendly slicing that generates proper code for
0-length result slices.
This is a takeover of Alexandru's original change,
(https://go-review.googlesource.com/#/c/12764/)
submittable now that the decompose phase is in.
Change-Id: I17d164cf42ed7839f84ca949c6ad3289269c9160
Reviewed-on: https://go-review.googlesource.com/13903
Reviewed-by: David Chase <drchase@google.com>
Basic ops, no particular optimization in the pattern
matching yet (e.g. x!=x for Nan detection, x cmp constant,
etc.)
Change-Id: I0043564081d6dc0eede876c4a9eb3c33cbd1521c
Reviewed-on: https://go-review.googlesource.com/13704
Reviewed-by: Keith Randall <khr@golang.org>
A comparison of the form l == r where l is an interface and r is
concrete performs a type assertion on l to convert it to r's type.
However, the compiler fails to zero the temporary where the result of
the type assertion is written, so if the type is a pointer type and a
stack scan occurs while in the type assertion, it may see an invalid
pointer on the stack.
Fix this by zeroing the temporary. This is equivalent to the fix for
type switches from c4092ac.
Fixes#12253.
Change-Id: Iaf205d456b856c056b317b4e888ce892f0c555b9
Reviewed-on: https://go-review.googlesource.com/13872
Reviewed-by: Russ Cox <rsc@golang.org>
MOVXload and MOVXstore opcodes have both an auxint offset
and an aux offset (a symbol name, like a local or arg or global).
Make sure we keep those values during rewrites.
Change-Id: Ic9fd61bf295b5d1457784c281079a4fb38f7ad3b
Reviewed-on: https://go-review.googlesource.com/13849
Reviewed-by: Josh Bleecher Snyder <josharian@gmail.com>
This CL takes a simple approach to spilling and loading flags.
We never spill. When a load is needed, we recalculate,
loading the arguments as needed.
This is simple and architecture-independent.
It is not very efficient, but as of this CL,
there are fewer than 200 flag spills during make.bash.
This was tested by manually reverting CLs 13813 and 13843,
causing SETcc, MOV, and LEA instructions to clobber flags,
which dramatically increases the number of flags spills.
With that done, all stdlib tests that used to pass
still pass.
For future reference, here are some other, more efficient
amd64-only schemes that we could adapt in the future if needed.
(1) Spill exactly the flags needed.
For example, if we know that the flags will be needed
by a SETcc or Jcc op later, we could use SETcc to
extract just the relevant flag. When needed,
we could use TESTB and change the op to JNE/SETNE.
(Alternatively, we could leave the op unaltered
and prepare an appropriate CMPB instruction
to produce the desired flag.)
However, this requires separate handling for every
instruction that uses the flags register,
including (say) SBBQcarrymask.
We could enable this on an ad hoc basis for common cases
and fall back to recalculation for other cases.
(2) Spill all flags with PUSHF and POPF
This modifies SP, which the runtime won't like.
It also requires coordination with stackalloc to
make sure that we have a stack slot ready for use.
(3) Spill almost all flags with LAHF, SETO, and SAHF
See http://blog.freearrow.com/archives/396
for details. This would handle all the flags we currently
use. However, LAHF and SAHF are not universally available
and it requires arranging for AX to be free.
Change-Id: Ie36600fd8e807ef2bee83e2e2ae3685112a7f276
Reviewed-on: https://go-review.googlesource.com/13844
Reviewed-by: Keith Randall <khr@golang.org>
This further reduces the number of flags spills
during make.bash by about 50%.
Note that GetG is implemented by one or two MOVs,
which is why it does not clobber flags.
Change-Id: I6fede8c027b7dc340e00d1e15df1b87bf2b2d9ec
Reviewed-on: https://go-review.googlesource.com/13843
Reviewed-by: Keith Randall <khr@golang.org>
