Jörn Horstmann e393f56d37 Improve autovectorization of to_lowercase / to_uppercase functions
Refactor the code in the `convert_while_ascii` helper function to make
it more suitable for auto-vectorization and also process the full ascii
prefix of the string. The generic case conversion logic will only be
invoked starting from the first non-ascii character.

The runtime on microbenchmarks with ascii-only inputs improves between
1.5x for short and 4x for long inputs on x86_64 and aarch64.

The new implementation also encapsulates all unsafe inside the
`convert_while_ascii` function.

Fixes #123712
2024-09-23 11:31:29 +02:00
..
2024-05-31 15:56:43 +10:00
2024-08-29 18:12:31 +08:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-09-09 19:39:43 -07:00
2024-09-09 19:39:43 -07:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-09-09 19:39:43 -07:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-04-11 21:42:35 -04:00
2024-04-22 18:48:47 +02:00
2024-04-11 21:42:35 -04:00
2024-02-24 19:49:16 +00:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-08-29 18:12:31 +08:00
2024-04-23 18:31:57 +02:00
2024-05-31 15:56:43 +10:00
2024-04-24 13:12:33 +01:00
2024-09-09 19:39:43 -07:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-06-16 17:19:25 +08:00
2024-05-31 15:56:43 +10:00
2024-05-03 14:32:08 +02:00
2024-05-31 15:56:43 +10:00
2024-07-14 13:48:29 +03:00
2024-05-31 15:56:43 +10:00
2024-04-22 18:48:47 +02:00
2024-07-14 13:48:29 +03:00
2024-09-09 13:29:47 +00:00
2024-05-31 15:56:43 +10:00
2024-07-14 13:48:29 +03:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-08-07 00:41:48 -04:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-05-31 15:56:43 +10:00
2024-09-09 19:39:43 -07:00
2024-03-29 15:37:43 -07:00
2024-02-23 18:04:01 -05:00
2024-06-19 21:26:48 +01:00
2024-05-31 15:56:43 +10:00
2024-09-09 19:39:43 -07:00

The files here use the LLVM FileCheck framework, documented at https://llvm.org/docs/CommandGuide/FileCheck.html.

One extension worth noting is the use of revisions as custom prefixes for FileCheck. If your codegen test has different behavior based on the chosen target or different compiler flags that you want to exercise, you can use a revisions annotation, like so:

// revisions: aaa bbb
// [bbb] compile-flags: --flags-for-bbb

After specifying those variations, you can write different expected, or explicitly unexpected output by using <prefix>-SAME: and <prefix>-NOT:, like so:

// CHECK: expected code
// aaa-SAME: emitted-only-for-aaa
// aaa-NOT:                        emitted-only-for-bbb
// bbb-NOT:  emitted-only-for-aaa
// bbb-SAME:                       emitted-only-for-bbb