From: Victor Stinner <vstinner@python.org>
Date: Mon, 3 Jun 2024 06:45:20 +0000 (+0200)
Subject: gh-119396: Optimize unicode_decode_utf8_writer() (#119957)
X-Git-Tag: v3.14.0a1~1645
X-Git-Url: http://git.ipfire.org/gitweb.cgi?a=commitdiff_plain;h=3ea9b92086240b2f38a74c6945e7a723b480cefe;p=thirdparty%2FPython%2Fcpython.git

gh-119396: Optimize unicode_decode_utf8_writer() (#119957)

Optimize unicode_decode_utf8_writer()

Take the ascii_decode() fast-path even if dest is not aligned on
size_t bytes.
---

diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 12782754753e..53160f1799f2 100644
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -4702,8 +4702,9 @@ ascii_decode(const char *start, const char *end, Py_UCS1 *dest)
     const char *p = start;
 
 #if SIZEOF_SIZE_T <= SIZEOF_VOID_P
-    assert(_Py_IS_ALIGNED(dest, ALIGNOF_SIZE_T));
-    if (_Py_IS_ALIGNED(p, ALIGNOF_SIZE_T)) {
+    if (_Py_IS_ALIGNED(p, ALIGNOF_SIZE_T)
+        && _Py_IS_ALIGNED(dest, ALIGNOF_SIZE_T))
+    {
         /* Fast path, see in STRINGLIB(utf8_decode) for
            an explanation. */
         /* Help allocation */
@@ -4948,9 +4949,7 @@ unicode_decode_utf8_writer(_PyUnicodeWriter *writer,
     const char *end = s + size;
     Py_ssize_t decoded = 0;
     Py_UCS1 *dest = (Py_UCS1*)writer->data + writer->pos * writer->kind;
-    if (writer->kind == PyUnicode_1BYTE_KIND
-        && _Py_IS_ALIGNED(dest, ALIGNOF_SIZE_T))
-    {
+    if (writer->kind == PyUnicode_1BYTE_KIND) {
         decoded = ascii_decode(s, end, dest);
         writer->pos += decoded;