By default clang-format interprets function-like macro invocations as
function calls. E.g. the argument of `CARBON_KIND(llvm::ListSeparator*
sep)` is interpreted as an expression, meaning the `*` is an infix
binary operator, so it inserts a space before the `*`. This change
teaches clang-format that `CARBON_KIND(x)` and
`CARBON_ASSIGN_OR_RETURN(x)` rewrite to `x`, which is close enough to
the truth to enable it to format them correctly. See the [clang-format
docs](https://clang.llvm.org/docs/ClangFormatStyleOptions.html#macros)
for details.
Echoing what was added in #5608, updating existing uses. Unfortunately
there's divergent behavior for operators versus constructors, so keeping
the nolint on those.
Trying to figure out an easy way to debug semir in the prelude, #5703
removed an option to set `--exclude-dump-file-prefix` to empty. But,
this is probably an improvement over that flow... With this change, it's
possible to add `//@dump-sem-ir-file` to a specific prelude file, and
its full IR will be printed. Additionally, it becomes an option with the
default `--dump-sem-ir-ranges=only` to add `//@dump-sem-ir-file` and get
the full file's IR.
toolchain/check/testdata/builtins/char/basics.carbon and
toolchain/lower/testdata/builtins/char.carbon are probably the most
interesting tests here. The parse tests is required because this adds a
new node kind, and we need coverage of it; but the attached info is
minor. There's a fair amount of test churn here because I'm adding the
Core.Char and Core.CharLiteral types as new singletons.
My intent here is that `CharId` is always a unicode code point, even
when the type is a `Char` and thus must be a single UTF-8 code unit
(single byte). This mainly means the stored value of a `CharValue` can
be printed internally without knowing the type.
---------
Co-authored-by: josh11b <15258583+josh11b@users.noreply.github.com>
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
Adapts `StringLiteral` to lex characters. Adds a `CharLiteral` token,
which contains a `CharLiteralValue` which is a straight unicode code
point (suggested by zygoloid).
---------
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
Also document why these are expected on `CARBON_KIND`. In
`kind_switch_test.cpp`, drop the `str` variable.
My recollection of the original discussion of `CARBON_KIND` is that it
should always have braces due to the risk of confusion for statement
interpretation, similar to a typical `if`/`else` but more subtle due to
the macro.
For example:
```
case CARBON_KIND(int n):
str << "int = " << n;
return str.TakeStr();
```
is equivalent to:
```
case CARBON_KIND(int n): {
str << "int = " << n;
}
return str.TakeStr();
```
This happens to work in context because `str` isn't scoped, but a
trivial refactoring to move `RawStringOstream str;` the first statement
of the `case` would probably have non-obvious results. For example:
```
case CARBON_KIND(int n):
RawStringOstream str; // Valid name shadowing, destructed without use.
str << "int = " << n; // Name lookup error on `n`.
return str.TakeStr();
```
Versus #5823, this is manually updated (still verified with `pre-commit
run -a`).
I also looked at updating prettier; adding a note why I'm not doing
that.
This is reducing ValueStore inference of types from `using`, and removes
`using ValueType = ...` from affected id types.
I'm adding a number of `using FooStore = ValueStore<FooId, Foo>` because
I think it's a little repetitive otherwise; often 4 cases where I'm
doing this: getter, const getter, member, and getter on `Context`. Note
we also have a number of `-> decltype(auto)` that were added I think
mainly to avoid repeating the type, but I'm not sure whether there'll be
agreement on replacing those and so am not changing them here.
I'm placing these aliases with the value type in general, because I
think it's probably easier to view that way. An alternative would be to
put all the types on `File`, but:
- That would be inconsistent with things like `InstStore`, which are
very `ValueStore`-adjacent and put with their value type.
- `File` would have a _lot_ of using's, and the accessors are already
noisy -- I think it would just make the file harder to skim.
Note this is the heart of what I'd brought up [on
Discord](https://discord.com/channels/655572317891461132/655578254970716160/1388199282250613019).
This PR still leaves CanonicalValueStore and BlockValueStore as things
to also add parameters to, but I thought it best to try breaking the set
of changes apart by type. Both of those rely on ValueStore, so
ValueStore needs to change first.
By moving dumping, we can have dumping occur before verification that
might CHECK-fail (e.g. parse tree and llvm IR verification).
I'm dropping vlogging of raw semir. It was only done when dumping, so
`-v` would print zero copies and `-v --dump-raw-sem-ir` would print two
copies. The lack of complaints about this suggests it's not needed.
I'm making a small change to drop newlines between textual and raw
semir. This is an edge case so I don't expect people to really notice in
general, but it seemed unusually aware of what's on a stream, and it
made it harder to do the dump_stream/raw_dump_stream approach, which I
felt would be decent in general, since check is the only phase that can
emit two different things (which I could also just drop -- we don't
really use raw semir anymore, it doesn't seem like a big need to be able
to print it with textual semir, but I'm assuming to just maintain
existing behavior).
In parse, we were previously dumping the tree on verification errors.
I'm removing that because now `--dump-parse-tree` should work fine,
where previously it wouldn't.
I've been mulling this mainly for the parameter complexity of
check/lower, but doing lex/parse for symmetry.
I'm motivated by the plan to move dumping for all of them into the
respective functions, because of discussion about llvm-verifier. That
basically would add another bool parameter (or more) to each of these.
My instinct is we're going to probably accrue a little more over time,
so I'm suggesting this as maybe adding the boundary a little simpler
and/or easier to read.
Note it may make sense to refactor a little further, e.g. maybe
Lower::Context could receive the full set of options and pick out what
it wants, but I figured creating the struct itself would be a decent
start.
I'm trying to put things into options when we can produce a reasonable
default if the user doesn't assign a value. I'm using an explicit
constructor so that values can be added without affecting every caller.
A different factoring would be to pass in everything through the param
struct, but that just felt weird when I was trying it out.
Removing `inst_namer` and `module_name` from `LowerToLLVM` params --
both of these can be inferred from `sem_ir`, and I'm not seeing a
particular reason to maintain them at the call site.
Split out `TokenInfo` to be able to easily write `using ValueType =
TokenInfo;` on `TokenIndex`. Also fixes a small type issue on
`ValueStore` that affected `mapped_iterator` behavior when writing
`old_tokens_it->first < next_offset`.
Moves LineInfo and CommentData out so that they can easily be set as
`ValueType` on the Index types. I've also been thinking about letting
`ValueStore` take `ValueType` as a parameter instead of requiring it to
be inferred this way, but for these it feels more consistent with the
rest of the toolchain to do it this way.
I'm not doing similar with `TokenInfo` just because the recovery token
splicing makes it more difficult to use `ValueStore`.
Teach CARBON_KIND_SWITCH to handle mutable lvalues and rvalues, and
CARBON_KIND to forward along rvalues so that it's possible to write
`case CARBON_KIND(const T& t)`, `case CARBON_KIND(T& t)`, and `case
CARBON_KIND(T&& t)`, depending on the type that was passed to
CARBON_KIND_SWITCH.
Replace all uses of VariantMatch with their equivalent of a switch using
CARBON_KIND_SWITCH, and remove the VariantMatch helper from the
codebase.
#5445 updates to bazel 8.2.1, this does more updates (including to
buildifier, which does autofixes like the `sh_test` loads in the other
PR).
Note I'm using the latest available clang-format wheel. That's not
really something I expect people to have installed, but should mostly be
consistent. I'm specifically skipping clang-format 18 because it had
some broad regressions, and 19 got really confused by a `requires` on a
trailing return. Using the latest seemed probably okay since most people
won't see the difference. Do note that trailing returns in macros,
https://github.com/llvm/llvm-project/issues/47664, seems to be cropping
up again as an issue.
The current behavior hits UBSAN and ASAN issues.
Note, `RequiresNullTerminator` is already set to `false` in
`source_buffer.cpp`; setting it in `compile_helper.cpp` is making things
more consistent. The related logic is an [assert
fail](https://github.com/llvm/llvm-project/blob/main/llvm/lib/Support/MemoryBuffer.cpp#L52).
This was fuzzer-discovered.
This consolidates Lex::TokenizedBuffer::DumpSemIRRange and
Parse::TreeAndSubtrees::TokenRange into a single InclusiveTokenRange,
also making the OverlapsWithDumpSemIRRange function take the new struct.
I considered switching to `llvm::iterator_range<Lex::TokenIterator>`,
but we often want to see if the range is size one. Using `TokenIterator`
just looked like it'd add a bunch of offsetting to make it work; I view
that as low-value overhead.
For example:
```
Lex::InclusiveTokenRange token_range = GetSubtreeTokenRange(node_id);
auto begin_loc = tree_->tokens().TokenToDiagnosticLoc(token_range.begin);
if (token_range.begin == token_range.end) {
return begin_loc;
}
auto end_loc = tree_->tokens().TokenToDiagnosticLoc(token_range.end);
```
would become:
```
llvm::iterator_range<Lex::TokenIterator> token_range = GetSubtreeTokenRange(node_id);
auto begin_loc = tree_->tokens().TokenToDiagnosticLoc(*token_range.begin());
if (token_range.begin() + 1 == token_range.end()) {
return begin_loc;
}
auto end_loc = tree_->tokens().TokenToDiagnosticLoc(*(token_range.end() - 1));
```
So I'm keeping the bespoke struct.
Syntax rationale is on `DumpSemIRRange` to try and record this, since
I'm not sure this belongs in the language design. The intent of this is
to be able to subset SemIR, which will be done separately in the
formatter.
---------
Co-authored-by: David Blaikie <dblaikie@gmail.com>
What this really does is avoids shadowing names, so that we can
comfortable have things like `Check::DiagnosticEmitter` or
`Check::DiagnosticLoc` without shadowing being a concern.
Note, down this path I'm also thinking about:
- Renaming misc DiagnosticConsumer/DiagnosticEmitter classes, possibly
just to DiagnosticConsumer/DiagnosticEmitter (so
`Check::DiagnosticEmitter` instead of `SemIRLocDiagnosticEmitter`).
- Dropping `Diagnostic` from `Emitter::DiagnosticBuilder`.
- But not for `Check::DiagnosticBuilder`, because `Check::Builder` would
be ambiguous.
- Renaming diagnostics/diagnostic_* to drop "diagnostic".
[Discussion about SemIRLoc ->
DiagnosticLoc](https://discord.com/channels/655572317891461132/655578254970716160/1353771570463768698)
reminded me of this (in particular the older [Check::DiagnosticBuilder
discussion](https://discord.com/channels/655572317891461132/655578254970716160/1344363562608627763)),
but I'd only do that rename if there's matching consensus about a path
forward where we keep SemIRLoc, and in a way that it's only ever used
for diagnostics (the divergence from which is at the root of current
LocId discussion).
I'm trying to keep that separate from a namespace addition for clarity.
We had a long discussion of this, so trying to document what seems to be
the conclusion... and also clean up the exceptions that I could find.
---------
Co-authored-by: Dana Jansens <danakj@orodu.net>
In line with the proposal in #4682, this changes the array syntax to be
array(T, N). `array` is a builtin keyword which must be followed by
parens containing two expressions and a separating comma.
The array type expression is still fully builtin, it does not forward to
a Core.Array library type yet. It merely adds the `ArrayType`
instruction, as was done with the previous syntax.
Followup work will change the instruction to reference to Core.Array,
once the library type exists and can be used directly.
---------
Co-authored-by: zygoloid <richard@metafoo.co.uk>
At present, we typically define a DiagnosticConverter, then store an
instance of it and a DiagnosticEmitter that wraps it. This is relatively
minor in general, but I've been trying to create more self-contained
DiagnosticEmitter classes (which hold their own DiagnosticConverter,
similar to NullDiagnosticEmitter), and there it just gets in the way.
Since we don't reuse DiagnosticConverter instances, this combines the
definition into DiagnosticEmitter. Mainly this means we don't have a
separate object in play, and less to carry around.
The most impact is probably to SemIRDiagnosticConverter, which was also
the most complex. Now `SemIRLocDiagnosticEmitter`, this gets some
different construction flow. Note in the PR I've split the file rename
to its own commit, to try to help delta views. However, the most
substantial parts of the refactoring are split into #4876, which this
depends upon.
High level, replacing `Id::Invalid` with `Id::None` and `Id::is_valid`
with `Id::has_value` for clarity, as discussed
[here](https://discord.com/channels/655572317891461132/655578254970716160/1331664574545395794).
The `IntId` refactoring is needed together with `AnyIdBase` because it's
also used with `ValueStore`.
Note, trying to be careful not to rewrite `EnumBase::InvalidIndex`, or
`is_valid` in general (e.g., `IdKind::is_valid`).
I've tried to sequence commits here:
1. Automatic replacements:
- `((?:Id|Index)(?: |::|\(|Base(?:\(|::)))Invalid((?:Index)?\W)` ->
`$1None$2`
- `<invalid>` -> `<none>`
- `InvalidNodeId` -> `NoneNodeId`
- `/\*invalid\*/` -> `/*none*/`
- `id((?:_|\(\))(?:\.|->))is_valid` -> `id$1has_value`
2. Manual edits:
- In `int.h` and `int_test.cpp`
- `IntT` has `is_value`, which I'm renaming to `is_embedded_value`.
- Manual edits to comments in this file.
- `AnyIdBase` and `IdBase`
- Declaration of `is_valid` -> `has_value`, `InvalidIndex` ->
`NoneIndex`.
- In `ids.h` and `ids.cpp`
- `is_valid` -> `has_value`
- `// An explicitly invalid ID.` -> `// An ID with no value.`; similar
for index
- Various math on `InvalidIndex` -> `NoneIndex`
- Various mentions of "valid" in comments
- In `value_store.h`, for `IdT::Invalid`, plus one comment
- In `impl.h` and `tokenized_buffer.h`, we had different initialization
of `::None` values (versus `ids.h` syntax) that I fixed manually.
- Spot checks to compile
- Particularly where `is_valid` replacements didn't catch spots due to
different naming.
3. Autoupdate tests
4. verbose.carbon (NOAUTOUPDATE)
5. Comment spot checks
Note there are probably other mentions of "Invalid" that should be swept
up, but I'd like to argue for merging and separating out remaining
cleanup since this is so sweeping (and likely to hit merge conflicts
from churn). We'll probably have lingering mentions of "invalid" for a
bit regardless, just because there are uses of "invalid" in non-Id APIs.
Disables three new warnings because they lean more towards style
conflicts than fixes. I've brought these up on #style.
Other than that, mostly fixing basic issues, and things that
clang-tidy-20 seems to fire where clang-tiday-16 didn't. One particular
curious case is `llvm::StringLiteral::data()` uses, which are flagged as
not strictly null-terminated; I'm switching to `const char*` in those
spots which matches `llvm::formatv`'s format argument, but feels worse.
I'm removing `run_clang_tidy.py` here because I'm observing it give
fewer warnings than `bazel build --config=clang-tidy -k
//toolchain/...`. The latter matches how we enforce in GitHub actions
(and also caches results, and suppresses output for files that have no
issues), so I'm dropping the bespoke script.
Just a small packing optimization. We currently have 222 `NodeKinds`, so
this reduces us to just 30ish more we can add without needing to pack
more. However, if we did, there would be a couple options for bringing
the count down by reusing `NodeKinds` and disambiguating based on the
token kind (the 29 infix operators as an example). Or we could just undo
this.
I'm expecting this to yield a small improvement. I'll see if I can get
better numbers since my machine's not really reliable, but here are some
basic values.
Also suggesting to draw the use of `::RawEnumType` for `TokenKind`,
since bit packing appears to work without it. Hoping the `static_assert`
is easier for people to understand the size of the field.
With the change:
```
----------------------------------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations Bytes Lines Tokens
----------------------------------------------------------------------------------------------------------------------------
BM_CompileAPIFileDenseDecls<Phase::Parse>/256 50399 ns 50359 ns 14336 104.588M/s 3.87217M/s 21.8629M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/1024 237823 ns 237629 ns 3072 136.721M/s 4.11986M/s 24.2058M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/4096 997645 ns 996771 ns 768 142.343M/s 4.04105M/s 23.9363M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/16384 4020308 ns 4018319 ns 192 152.041M/s 4.05966M/s 24.0874M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/65536 16691390 ns 16683058 ns 48 151.317M/s 3.92374M/s 23.2936M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/262144 75265735 ns 75233476 ns 8 135.842M/s 3.48421M/s 20.6862M/s
```
Without the change:
```
----------------------------------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations Bytes Lines Tokens
----------------------------------------------------------------------------------------------------------------------------
BM_CompileAPIFileDenseDecls<Phase::Parse>/256 51515 ns 51480 ns 13312 102.312M/s 3.78789M/s 21.387M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/1024 241040 ns 240900 ns 3072 134.865M/s 4.06392M/s 23.8771M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/4096 985593 ns 984657 ns 768 144.094M/s 4.09077M/s 24.2308M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/16384 4109327 ns 4105496 ns 192 148.813M/s 3.97345M/s 23.576M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/65536 17459655 ns 17446006 ns 48 144.7M/s 3.75215M/s 22.275M/s
BM_CompileAPIFileDenseDecls<Phase::Parse>/262144 80802815 ns 80737489 ns 8 126.581M/s 3.24668M/s 19.276M/s
```
---------
Co-authored-by: Chandler Carruth <chandlerc@gmail.com>
This is essentially the result of looking at `.begin()` uses. We also
frequently do `std::shuffle`, but unfortunately STLExtras doesn't
provide a wrapper for that.
The first change here is to canonicalize away bit width when tracking
integers in our shared value store. This lets us have a more definitive
model of "what is the mathematical value". It also frees us to use more
efficient bit widths when available, such as bits inside the ID itself.
For canonicalizing, we try to minimize the width adjustments and
maximize the use of the SSO in APInt, and so we never shrink belowe
64-bits and grow in multiples of the word bit width in the
implementation. We also canonicalize to the signed 2s compliment
representation so we can represent negative numbers in an intuitive way.
The canonicalizing requires getting the bit width out of the type and
adjusting to it within the toolchain when doing any kind of math, and
this PR updates various places to do that, as well as adding some
convenience APIs to assist.
Then we take advantage of the canonical form and embed small integers
into the ID itself rather than allocating storage for them and
referencing them with an index. This is especially helpful for the
pervasive small integers such as the sizes of types, arrays, etc. Those
no longer require indirection at all. Various short-cut APIs to take
advantage of this have also been added.
This PR improves lexing by about 5% when there are lots of `i32` types.
---------
Co-authored-by: Dana Jansens <danakj@orodu.net>
Co-authored-by: Carbon Infra Bot <carbon-external-infra@google.com>
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
This is in anticipation of making the integer value store be customized
heavily. I'd like to extract it from the common code when doing that, so
first disentangling them here without any intended change in
functionality or behavior to enable that.
I've tried to update `#include`s to be as minimal as I can and added a
few missing includes spotted in the process.
I've split the test for value store to include what was easy focused on
just the value store templates rather than the unified shared value
stores.
This might surface some opportunities for adding more tests, but for
this PR, just doing the minimal restructuring.
This rejects type literals with more digits than we can lex without
APInt's help, and using a custom diagnostic. This is a pretty arbitrary
implementation limit, I'm wide open to even more strict rules here.
Despite no special casing and a very simplistic approach, by not using
APInt this completely eliminates the lexing overhead for `i32` in the
generated compilation benchmark where that specific type literal is very
common. We see a 10% improvement in lexing there:
```
BM_CompileAPIFileDenseDecls<Phase::Lex>/256 39.0µs ± 4% 34.8µs ± 2% -10.86% (p=0.000 n=19+20)
BM_CompileAPIFileDenseDecls<Phase::Lex>/1024 180µs ± 1% 158µs ± 2% -12.22% (p=0.000 n=18+20)
BM_CompileAPIFileDenseDecls<Phase::Lex>/4096 731µs ± 2% 641µs ± 1% -12.31% (p=0.000 n=18+20)
BM_CompileAPIFileDenseDecls<Phase::Lex>/16384 3.20ms ± 2% 2.86ms ± 2% -10.47% (p=0.000 n=18+19)
BM_CompileAPIFileDenseDecls<Phase::Lex>/65536 13.8ms ± 1% 12.4ms ± 2% -9.78% (p=0.000 n=18+19)
BM_CompileAPIFileDenseDecls<Phase::Lex>/262144 64.0ms ± 2% 58.4ms ± 2% -8.70% (p=0.000 n=19+18)
```
This starts to fix a TODO in the diagnostic for these by giving a
reasonably good diagnostic about a very large type literal. However, in
practice it regresses the diagnostics because error tokens produce noisy
extraneous diagnostics from parse and check currently. Leaving the TODO
there, and I have a follow-up PR to start improving the extraneous
diagnostics.
Previously in an optimized build we'd produce bogus tokens, such as
tokens with incorrect IdentifierIds, and in a debug build we would try
to CHECK-fail -- but actually wouldn't, because we're incorrectly
checking for `2 << bits` instead of `1 << bits`. I hit this while I was
trying to do some profiling and was seeing some very strange
diagnostics.
The diagnostic is pointed at the first token that is beyond the limit to
help people determine where to split their files.
---------
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
This is improving the comment production to produce fewer distinct
comments.
At present, comment processing uses strict prefix matching. It either
expects `// ` (with a space) for valid comments, or just `//` (without a
space) for invalid comments that lacked the space.
As a consequence, the following would be three comments:
```
// Comment 1
//
//
// Comment 4
```
This is because a 3-character prefix is used for valid comments. The
prefix switches between lines 1 and 2, and again between lines 3 and 4,
each resulting in a separate comment.
For contrast, this is one comment because only a 2-character prefix is
used:
```
//Comment 1
//
//
//Comment 4
```
That's because all lines lack a suffix space.
Additionally, with SIMD 16-byte boundaries, further splits can occur if
processing needs to transition to non-SIMD.
Here, I'm trying to just address all of this by:
1. Stitching together adjacent comments. Since a lexed comment starts at
the `//` excluding the indent, the delta from the prior comment must be
precisely the indent.
2. Adding support for switching from SIMD to non-SIMD on file
boundaries.
I considered trying to have a separate `//\n` prefix for SIMD processing
of `// `, but I wasn't sure about the tradeoff of doing both at the same
time (in particular, it'd require constructing a string for the
different prefix), thus this stitch approach. This does mean multiple
passes will be required for a typical long comment structure using blank
comment lines to separate paragraphs (for performance reasons, I will
recommend engineers not write comm... nevermind).
---------
Co-authored-by: Chandler Carruth <chandlerc@gmail.com>
In order to format comments, it's helpful if they're tracked. This
tracks them separately from tokens in order to avoid interfering with
parse; it'd be inconvenient if comment tokens could show up in arbitrary
locations, albeit possible to support.
This additionally extracts out the TokenIterator support into a template
in order to generally have it available for IndexBase types. I'm only
adding it for CommentInfo, not sure if we'll want it elsewhere, but this
structure still felt like a good fit.
This is a primarily automated change:
- Search & replace for capitalization
-
`(CARBON_DIAGNOSTIC\((?:\n\s+)?\w+,(?:\n\s+)?\s\w+,(?:\n\s+)?\s")([A-Z])`
- `$1\L$2`
- Search & replace for period
-
`(CARBON_DIAGNOSTIC\((?:\n\s+)?\w+,(?:\n\s+)?\s\w+,(?:\n\s+)?\s"(?:[^)]|\n)+)\.("[,)])`
- `$1$2`
- Limited search & replace for `ERROR: ` -> `error: ` in streamed things
- Leaving a TODO for command_line because there's more cleanup that can
be done there
- Modify diagnostic_consumer.cpp
- ERROR -> error
- WARNING -> warning
---------
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
This switches `DCHECK` and `FATAL` as well.
The goal is to reduce the code size impact of these assertions so that
we can keep more of them enabled. Currently, the largest cost I see from
`CHECK` is not the actual check or the cold code itself, but actually
the failure to inline trivial functions due to the presence of the cold
code. This means that our goal isn't to reduce apparent code size in the
final binary but the LLVM IR cost assessed for these routines in the
inliner, which closely correlates with code size but is a bit different.
As discussed in #4283, experimentation shows that a single function call
with a minimal number of arguments is the lowest cost model for these.
This is easily achieved with a format-string API that internally uses
`llvm::formatv`. This PR is essentially the `CHECK` version of #4283.
However, the check macros are substantially harder to make work with
both format strings and streaming because they also take a condition.
Also, unexpectedly, I was very successful at devising a regular
expression based automated rewrite from the streaming to the format
string form with only low 10s of manual fixes. This includes compacting
strings broken up across lines, etc. Given how well that went, I've
prepared this PR which just directly switches to the format string API
and migrate everything to use it.
One nice side-effect is that the format string approach ends up greatly
simplifying the implementation here as well.
This is ... *shockingly* effective. Parsing speeds up by more than 3%
with just this change. And checking speeds up by **8%** with this change
alone:
```
BM_CompileAPIFileDenseDecls<Phase::Parse>/256 86.3µs ± 1% 82.9µs ± 1% -3.94% (p=0.000 n=17+19)
BM_CompileAPIFileDenseDecls<Phase::Parse>/1024 431µs ± 1% 415µs ± 1% -3.76% (p=0.000 n=18+19)
BM_CompileAPIFileDenseDecls<Phase::Parse>/4096 1.77ms ± 1% 1.71ms ± 1% -3.18% (p=0.000 n=18+19)
BM_CompileAPIFileDenseDecls<Phase::Parse>/16384 7.44ms ± 1% 7.17ms ± 2% -3.56% (p=0.000 n=18+20)
BM_CompileAPIFileDenseDecls<Phase::Parse>/65536 30.7ms ± 1% 29.7ms ± 1% -3.15% (p=0.000 n=18+20)
BM_CompileAPIFileDenseDecls<Phase::Parse>/262144 131ms ± 1% 127ms ± 1% -2.81% (p=0.000 n=18+18)
BM_CompileAPIFileDenseDecls<Phase::Check>/256 878µs ± 2% 800µs ± 1% -8.91% (p=0.000 n=19+20)
BM_CompileAPIFileDenseDecls<Phase::Check>/1024 1.88ms ± 2% 1.72ms ± 1% -8.56% (p=0.000 n=19+20)
BM_CompileAPIFileDenseDecls<Phase::Check>/4096 5.78ms ± 2% 5.28ms ± 1% -8.70% (p=0.000 n=20+18)
BM_CompileAPIFileDenseDecls<Phase::Check>/16384 21.9ms ± 1% 20.1ms ± 1% -8.02% (p=0.000 n=18+20)
BM_CompileAPIFileDenseDecls<Phase::Check>/65536 90.4ms ± 2% 83.1ms ± 1% -8.04% (p=0.000 n=19+20)
BM_CompileAPIFileDenseDecls<Phase::Check>/262144 381ms ± 2% 352ms ± 1% -7.79% (p=0.000 n=19+19)
```
---------
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
Co-authored-by: josh11b <15258583+josh11b@users.noreply.github.com>
This makes each token info consist of 8 bytes of data:
- 1 byte of the kind
- 1 bit for whitespace tracking
- 23 bits of payload
- 32 bits for byte offset in the file
This builds directly on representing the location of the token as
a single 32-bit offset, now compressing the rest of the data into
a single 32-bit bitfield.
This adds some implementation limits: we can no longer lex more than
2^23 tokens in a single source file. Nor can we have more than 2^23
string literals, integer literals, real literals, or identifiers. Only
the first of these is even close to an issue, and even then seems
unlikely to ever be a problem in practice.
The memory efficiency here is great and the motivating goal. But to make
this work well, we also need to streamline how we create the tokens.
Otherwise, all the bit fiddling can end up erasing our gains. This PR
adds a number of APIs to manage creating and accessing the now
significantly more complex storage of token infos to try and help with
this.
One big change required to simplify the writes here is to switch from
computing whether a token has trailing space after-the-fact to
pre-computing whether a token will have leading space. That lets us have
the leading space information available immediately when forming the
token, and avoids doing a single bit flip afterward.
Another change that helps with this representation is to minimize the
updating of groups after-the-fact. The code now tries to set the opening
index directly when creating the closing token and only updates the
opening group afterward. Because of the bit packing, this is a reduction
of 0.5% of dynamic instructions in the compile benchmark, and has
dramatic improvements for the grouping symbol focused benchmarks.
All combined, this is a significant improvement on the lexer-focused
benchmarks despite the added complexity, and a significant win on our
compile time benchmarks due to both the lexer improvements and
downstream memory density improvements: 5-12% reduction in lex time,
growing larger as files get larger. About a 4.5% reduction in parse
time, and even a 1-2% reduction in total check time. =D
---------
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
Co-authored-by: Geoff Romer <gromer@google.com>
First, this replaces the separate line index and column index in the
token information with a single 32-bit byte offset of the token. This is
then used to compute line and column numbers with a binary search of the
line structure and then using that to compute the column within the
line. In practice, this is _much_ more efficient:
- Smaller token data structure. This will hopefully combine with a
subsequent optimization PR that shrinks the token data structure still
further.
- Fewer stores to form each token's information in the tight hot loop of
the lexer.
- Less state to maintain while lexing, fewer computations while lexing.
We only have to search to build the line and column information off the
hot lexing path, and so this ends up being a significant win and shrinks
some of the more significant data structures.
Second, this shrinks the line start to a 32-bit integer and removes the
line length. Our source buffer already ensures we only have 2 GiB of
source with a nice diagnostic. I've just added a check to help document
this in the lexer. The line length can be avoided in all of the cases it
was being used, largely by looking at the next line's start and working
from there. This also precipitated cleaning up some code that dated from
when lines were only built during lexing rather than being pre-built,
which resulted in nice simplifications.
With this PR, I think it makes sense to re-name a bunch of methods on
`TokenizedBuffer`, but to an extent that was already needed as these
methods somewhat predate the more pervasive style conventions. I avoided
that here to keep this PR focused on the implementation change, I'll
create a subsequent PR to update the API to both better nomenclature and
remove deviations from our conventions.
There may also be a way to de-duplicate the binary search in the
diagnostic location conversion and the main line accessor binary search,
but it wasn't obvious to me that it would be a net savings, so left it
alone for now.
The performance impact of this varies quite a bit...
The lexer's benchmark improves pretty consistent across the board on
both x86 and Arm. For x86, where I have nice comparison tools, it
appears 3% to 20% faster depending on the specific pattern. For Arm
server CPUs at least it seems a much smaller but still an improvement.
The overall compilation benchmarks however don't improve much with these
changes alone on x86. Significant reduction in instruction count
required for lexing, but the overall performance is bottlenecked
elsewhere in the overall compilation it seems. However, on Arm, despite
the more modest gains in special cases of lexing, this shows fairly
consistent 1-2% improvements in overall lexing performance on our
compilation benchmark. And the expectaiton is these improvements will
compound with subsequent work to further compact our representation.
---------
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
This uses a heuristic reserve to greatly reduce hashtable growth of the
identifiers hashtable. The design of the hashtable itself is optimized
around compact memory use and is especially slow to grow and so this has
an outsized impact.
The heuristic was computed using `scripts/source_stats.py` and looking
at C++ codebases. We may want to periodically re-evaluate it as Carbon
code emerges and we have better data on its distributions of tokens.
This also required fixing the `Reserve` method on `CanonicalValueStore`
that wasn't actually used anywhere and so didn't even compile correctly.
I added it to the relevant unit test so it is at least compiled locally
to its definition.
It turns out we can make these work with very minimal complexity because
the LF is still in the right place either way. This also lets us easily
support mixtures of LF and CR+LF line endings gracefully. We create the
line structures around the LF bytes and then have the byte-dispatch loop
notice a CR followed by an LF and skip to the LF behavior.
Rather than add the remaining complexity around supporting bare CR and
LF+CR sequences (both of which are quite rare now), this just adds
diagnostics when we encounter a CR byte that won't fall out of our CR+LF
handling. This is a better experience for users than the alternative. We
still have a TODO to handle the full complexity of vertical whitespace,
but I've updated it to reflect that the common case should be handled
already.
This isn't complete though: we need to add support in string literal
lexing, and we need to teach the diagnostic rendering to handle the
error messages above better. But those will be future PRs, this is
enough to unblock folks who happen to edit a Carbon source file with
notepad on Windows which seems important.
---------
Co-authored-by: Richard Smith <richard@metafoo.co.uk>
Just spotted these while looking at warnings that seem to fire on our
code are probably are things we'd fix if we saw them. None of these seem
important FWIW.
Also removes a redundant flag that is part of `-Wall`.
I have a follow-up for the high-value warning I spotted that motivated
me to look at all of this. But it's noisy so kept it as a separate PR.
The expected parse tree size is validated for valid parse trees, which
is possible even if lex encounters an error. In that case, we were
missing recovery tokens, resulting in a crash. AddToken should be the
only place we call `token_infos_.push_back`.
Instead of special-casing tokens starting with `r`, lex them as normal
identifiers, and add a special case to `#` handling to detect if the
previous token was an `r` identifier.
This roughly doubles the time to lex a raw identifier, because we do two
hash table insertions rather than one, and probably slightly slows down
lexing string literals starting with `#`, but should remove the 2%
overhead to identifier lexing from the previous approach.
In parse, form a list of methods that are defined inline, tracking where
they start, where they end, and which other inline methods are nested
within them.
In check, when we reach an inline method body, skip it and add it to a
worklist to be processed later. We also track when we reach the start
and end of a context in which inline method bodies are deferred, so that
we know when to replay the bodies.
When suspending a function definition to be processed later, the
`DeclNameStack` entry is moved to separate storage, including popping
the corresponding scopes from the scope stack and removing the
corresponding lexical names from lexical lookup. Later, when we return
to the function and parse its definition, the `DeclNameStack` entry is
restored. The same is done when we reach the end of a nested context
that can have inline methods, so that we can reenter the nested scope
before processing its members.
---------
Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
Since the addition of TranslateArg, I don't think this type is going to
go away (cutting a TODO). Refactoring names slightly to fit the current
role, and adding const to ConvertLocation.
Richard Smith
Geoff Romer
Jon Ross-Perkins
Dana Jansens
Thomas Köppe
Boaz Brickner
