[php-src] PR #21726: ext/uri: speed up Uri\Rfc3986\Uri component reads (~22%)
| From: | iliaal | Date: | Sat, 11 Apr 2026 19:59:32 +0000 |
| Subject: | [php-src] PR #21726: ext/uri: speed up Uri\Rfc3986\Uri component reads (~22%) | ||
| Groups: | php.git-pulls | ||
| Request: | Send a blank email to git-pulls+get-35678@lists.php.net to get a copy of this message | ||
Pull Request: https://github.com/php/php-src/pull/21726
Author: iliaal
## Summary
Speed up
Uri\Rfc3986\Uri parse + component reads by ~22% on a
realistic URL mix. Five targeted changes in ext/uri/uri_parser_rfc3986.c;
the full per-change breakdown is in the commit message.
## Benchmark
17 URL shapes: plain http/https, deep paths, with query/fragment, with
userinfo, IPv4, IPv6, mailto, URN, data, file, relative. 100K iterations
per run, 1.7M parses per benchmark, CPU pinned via taskset -c 0,
same-session A/B (stash + rebuild + rerun each direction).
| | baseline | optimized | delta |
|--|--|--|--|
| Uri\Rfc3986\Uri::parse only | 0.394s (4.30M/s) | 0.378s (4.50M/s) | −4% |
| Uri\Rfc3986\Uri::parse + 7 reads | **0.842s (2.02M/s)** | **0.653s (2.60M/s)** |
**−22% / +29% throughput** |
| Uri\WhatWg\Url::parse only | 0.410s (4.15M/s) | 0.408s (4.16M/s) | noise |
| Uri\WhatWg\Url::parse + 8 reads | 0.599s (2.84M/s) | 0.592s (2.89M/s) | noise |
The WhatWG rows are a control. This patch only touches the RFC3986
backend, so WhatWG throughput should stay within noise, and it does.
## What's in the patch
1. get_normalized_uri() aliases the raw URI when
uriNormalizeSyntaxMaskRequiredExA reports a zero dirty mask,
skipping the full uriCopyUriMmA + uriNormalizeSyntaxExMmA pass
for the common case where the parsed URI is already canonical. The
mask is cached so repeated non-raw reads on one instance scan once.
Biggest single contributor.
2. Port cache on the uris struct. Parse-time validation stashes the
converted zend_long; the first port_read serves from cache;
writes invalidate.
3. port_str_to_zend_long_checked swaps stack-copy + ZEND_STRTOUL
for an inline digit accumulator. Uriparser has already validated
the text as ASCII digits only, so strtoul's parse-flags, sign
handling, base detection, and stop-char scanning are dead weight.
4. uriparser_create_uris uses emalloc + targeted field init
instead of ecalloc, avoiding ~440 bytes of zeroing per parse.
5. php_uri_parser_rfc3986_destroy skips
uriFreeUriMembersMmA(&normalized_uri, ...) when it was never
built or when it aliases the raw uri. Required by the emalloc
change since the struct is no longer fully zeroed.