From: Mattias EngdegÄrd Date: Sat, 1 Apr 2023 08:53:50 +0000 (+0200) Subject: ; * src/fns.c: Use if instead of #ifdef X-Git-Url: http://git.eshelyaron.com/gitweb/?a=commitdiff_plain;h=f5891da70aae4deb465e8f193a172fbad0b4485f;p=emacs.git ; * src/fns.c: Use if instead of #ifdef * src/fns.c (HAVE_FAST_UNALIGNED_ACCESS, load_unaligned_size_t): Always define these. (Fstring_lessp): Use if instead of #ifdef. --- diff --git a/src/fns.c b/src/fns.c index 47def5c43b4..e92ef7e4c81 100644 --- a/src/fns.c +++ b/src/fns.c @@ -452,6 +452,9 @@ If string STR1 is greater, the value is a positive number N; || defined __s390__ || defined __s390x__) \ && defined __OPTIMIZE__ #define HAVE_FAST_UNALIGNED_ACCESS 1 +#else +#define HAVE_FAST_UNALIGNED_ACCESS 0 +#endif /* Load a word from a possibly unaligned address. */ static inline size_t @@ -462,8 +465,6 @@ load_unaligned_size_t (const void *p) return x; } -#endif - DEFUN ("string-lessp", Fstring_lessp, Sstring_lessp, 2, 2, 0, doc: /* Return non-nil if STRING1 is less than STRING2 in lexicographic order. Case is significant. @@ -504,18 +505,16 @@ Symbols are also allowed; their print names are used instead. */) /* String data is normally allocated with word alignment, but there are exceptions (notably pure strings) so we restrict the wordwise skipping to safe architectures. */ -#ifdef HAVE_FAST_UNALIGNED_ACCESS + if (HAVE_FAST_UNALIGNED_ACCESS) { /* First compare entire machine words. */ int ws = sizeof (size_t); const char *w1 = SSDATA (string1); const char *w2 = SSDATA (string2); - while (b < nb - ws + 1 - && (load_unaligned_size_t (w1 + b) - == load_unaligned_size_t (w2 + b))) + while (b < nb - ws + 1 && load_unaligned_size_t (w1 + b) + == load_unaligned_size_t (w2 + b)) b += ws; } -#endif /* Scan forward to the differing byte. */ while (b < nb && SREF (string1, b) == SREF (string2, b))