Issue #387 pointed out that integer float data is printed incorrectly, such
that it is not parsed correctly by the Go assembler. Specifically, integer
values need the decimal point, otherwise they will be treated as integers. For
example, 1 must be represented as `$(1.)` or `$(1.0)` to be parsed correctly.
This PR fixes that problem and adds a regression test. The root of the
problem was that the formatting verb `%#v` does not have the right behavior
for integers. We fix it by deferring to custom `String()` function for the
float operand types.
Fixes#387Closes#388
Commit 759be3dad9 bumped our Go
requirement to 1.18 which allows us to drop support for old-style
`+build` tags. This change runs `go fix ./...` to remove them, and
updates some remaining code generators that produced `+build` lines.
* Bump CI to Go 1.19
* Update golang/go edwards25519 test
* Apply formatting to printer stubs output (to get correct comment formatting)
* Bump gofumpt version
This fixes a bug in argument size calculation in the case where the function
has no return values. Previously it was padding the argument struct to max
alignment, but this only happens if there are return values following.
Updates #191
Currently `avo` uses `BP` as a standard general-purpose register. However, `BP` is used for the frame pointer and should be callee-save. Under some circumstances, the Go assembler will do this automatically, but not always. At the moment `avo` can produce code that clobbers the `BP` register. Since Go 1.16 this code will also fail a new `go vet` check.
This PR provides a (currently sub-optimal) fix for the issue. It introduces an `EnsureBasePointerCalleeSaved` pass which will check if the base pointer is written to by a function, and if so will artificially ensure that the function has a non-zero frame size. This will trigger the Go assembler to automatically save and restore the BP register.
In addition, we update the `asmdecl` tool to `asmvet`, which includes the `framepointer` vet check.
Updates #156
Adds a regression test based on klauspost/compress#186. This necessitated some related changes:
* Mark "RET" as a terminal instruction
* printer refactor to maintain compatibility with asmfmt
* Tweaks to other regression tests to ensure they are run correctly in CI
Updates #100#65#8
Issue #100 demonstrated that register allocation for aliased registers is
fundamentally broken. The root of the issue is that currently accesses to the
same virtual register with different masks are treated as different registers.
This PR takes a different approach:
* Liveness analysis is masked: we now properly consider which parts of a register are live
* Register allocation produces a mapping from virtual to physical ID, and aliasing is applied later
In addition, a new pass ZeroExtend32BitOutputs accounts for the fact that 32-bit writes in 64-bit mode should actually be treated as 64-bit writes (the result is zero-extended).
Closes#100
Adds support for a `CancellingInputs` instruction flag, to indicate cases like `XORQ R10, R10` where the instruction actually does not depend on the value of `R10` at all.
Closes#89
In some cases natural use of abstraction in avo programs can lead to redundant move instructions, specifically self-moves such as MOVQ AX, AX. This does not produce incorrect code but it is incorrect and inelegant.
This diff introduces a PruneSelfMoves pass that removes such unnecessary instructions.
Closes#76
The Go assembler merges MOVD/MOVQ instruction forms. The logic in the
avo instruction loader was discarding the MOVD forms. This diff should
merge them correctly.
Updates #50
