Previously Collect() was used for types that implemented
CollectMemUsage() but otherwise Add() was used. This required the caller
to think about the type of the field and know/decide which method to
use.
Now, the caller always uses Collect() unless they are adding specific
byte values, in which case Add is used. Typically then, Add will only be
used to implement the CollectMemUsage() function.
To do this we require all Collect() methods to be templates so that they
all be a single overload set. The Collect on BumpPtrAllocator is
converted to a template that checks
`std::same_as<llvm::BumpPtrAllocator, T>`.
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>
Move subtree sizes over to TreeAndSubtrees, using the different
structure to represent the additional parse work that occurs, as well as
making it clear which functions require the extra information. My intent
is to make it hard to use this by accident.
The subtree size is still tracked during Parse::Tree construction. I
think a lot of that can be cleaned up, although we use it during
placeholder assignment so it may take some work. I wanted to see what
people thought about this before taking action on such a change.
I'm using a 1m line source file generated by #4124 for testing. Command
is `time bazel-bin/toolchain/install/prefix_root/bin/carbon compile
--phase=check --dump-mem-usage ~/tmp/data.carbon`
At head, what I'm seeing is:
```
...
parse_tree_.node_impls_:
used_bytes: 61516116
reserved_bytes: 61516116
...
Total:
used_bytes: 447814230
reserved_bytes: 551663894
...
1.43s user 0.14s system 99% cpu 1.565 total
```
With `Tree::Verify` disabled completely, it looks like:
```
parse_tree_.node_impls_:
used_bytes: 41010744
reserved_bytes: 41010744
...
Total:
used_bytes: 427308858
reserved_bytes: 531158522
...
1.20s user 0.13s system 99% cpu 1.332 total
```
Re-enabling just the basic verification (what is now `Tree::Verify`),
I'm seeing maybe 0.05s slower, but that's within noise for my system. I
do see variability in my timing results, and overall I think this is a
0.2s +/- 0.1s improvement versus the earlier (always testing `Extract`
code) implementation. That's opt; debug builds will be unaffected,
because the same checking occurs as before.
Note, the subtree size is a third of the node representation, which is
why I'm showing the decrease in memory usage here.
Dana Jansens
Jon Ross-Perkins