zend_execute: Mark zend_get_executed_*() as __attribute__((pure))

[TimWolla]

These functions do not modify the state of the program and depend on thread-safe global variables only.

---

This is a possible precursor to #18995.

[TimWolla]

CI Benchmark indicates a slight slowdown. Can anyone with a “reliable CPU” do another non-valgrind check to see if this actually resulted in regressions? My CPU is notoriously bad at benchmarking.

[iluuu1994]

❯ b 50 --mode=cycles 'sapi/cli/php-old -n benchmark/repos/symfony-demo-2.2.3/public/index.php' 'sapi/cli/php-new -n benchmark/repos/symfony-demo-2.2.3/public/index.php'
Old:       1.694 G ± 991.360 k (0.059%) 
New:       1.696 G ± 1.041 M (0.061%)
Diff:      2.264 M (0.134%)
T-test:    5.679
P-value:   0.001

First test looks to be slower.

Edit: Second run approx. confirms.

❯ b 50 --mode=cycles 'sapi/cli/php-old -n benchmark/repos/symfony-demo-2.2.3/public/index.php' 'sapi/cli/php-new -n benchmark/repos/symfony-demo-2.2.3/public/index.php'
Old:       1.697 G ± 1.114 M (0.066%)   
New:       1.698 G ± 957.368 k (0.056%)
Diff:      1.337 M (0.079%)
T-test:    3.282
P-value:   0.008

[TimWolla]

[iluuu1994]

Can relate

[nielsdos]

Coming from your other PR, I see now why you submitted this.
I don't know why this worsens performance, but indeed the benchmark seems to indicate now a deviation higher than noise.
It could be many different reasons, from alignment of functions/loops to bad compiler spill/rematerialization decisions...

[iluuu1994]

Given that this PR seems to lead to a unclear performance degradation, I'd prefer not to merge it. We can investigate why it happens, and potentially create a bug report if we can identify a compiler-related culprit. It may also be alignment-related, as Niels mentioned. I have thought about a way to align the function offsets of two binaries when I encountered this issue myself, given that -falign-functions and friends didn't help. One way might be to build the binaries, inspect the elf-file and calculate the necessary nops to insert, and then use llvm-bolt's --pad-funcs-before flag. Not sure whether that will work. Suggestions are welcome.

[iluuu1994]

Also, for completion's sake: I tried to apply pure and const with an automated script last year:
master...iluuu1994:php-src:script-fix-const-pure-attributes

The entire PR showed a big degradation, although I don't remember by how much nor how I measured it. It would ofc be great if this unlocked some optimization opportunities. In that case, it might be better to go the automated route, rather than applying it bit-by-bit by hand.

[TimWolla]

I've rebased this one a final time and now the Symfony Benchmark indicates an improvement 🤷

[iluuu1994]

I will re-run my benchmarks later today.

[nielsdos]

Let's wait for Ilija's benchmark, but I'm in favour of doing this.

[iluuu1994]

Based on cycles counted by perf, I still see a significant regression on Symfony Demo:

// Two separate runs
Δ = 1.173 M (+0.223%, p < 0.001)
Δ = 1.040 M (+0.198%, p < 0.001)

And a significant improvement on Zend/bench.php:

// Two separate runs
Δ = -7.186 M (-0.649%, p < 0.001)
Δ = -8.979 M (-0.812%, p < 0.001)

And a huge regression on Zend/micro_bench.php:

// Two separate runs
Δ = 138.389 M (+2.754%, p < 0.001)
Δ = 120.192 M (+2.390%, p < 0.001)

It does seem like this is most likely layout related. It would be really great to invest some time into this idea to hopefully solve this once and for all. Otherwise we keep wasting hours with people benchmarking and following false leads, or potentially dropping good changes. Alternatively we might try to run LLVM bolt on both binaries, but it's hard to tell how fair this comparison will be. I have now posed this question on the LLVM #bolt Discord channel. If you have other ideas, please share.