From e7b2dd2e38ca595ea09f2bff8c0d8fc45e88cd4e Mon Sep 17 00:00:00 2001 From: "Richard M. Stallman" Date: Sun, 24 Sep 2006 20:28:49 +0000 Subject: [PATCH] (compile_pattern_1): Don't BLOCK_INPUT. --- src/search.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/src/search.c b/src/search.c index a74a320c356..e26b8f883e1 100644 --- a/src/search.c +++ b/src/search.c @@ -171,10 +171,11 @@ compile_pattern_1 (cp, pattern, translate, regp, posix, multibyte) cp->buf.multibyte = multibyte; cp->whitespace_regexp = Vsearch_spaces_regexp; cp->syntax_table = current_buffer->syntax_table; - /* Doing BLOCK_INPUT here has the effect that - the debugger won't run if an error occurs. - Why is BLOCK_INPUT needed here? */ - BLOCK_INPUT; + /* rms: I think BLOCK_INPUT is not needed here any more, + because regex.c defines malloc to call xmalloc. + Using BLOCK_INPUT here means the debugger won't run if an error occurs. + So let's turn it off. */ + /* BLOCK_INPUT; */ old = re_set_syntax (RE_SYNTAX_EMACS | (posix ? 0 : RE_NO_POSIX_BACKTRACKING)); @@ -187,7 +188,7 @@ compile_pattern_1 (cp, pattern, translate, regp, posix, multibyte) re_set_whitespace_regexp (NULL); re_set_syntax (old); - UNBLOCK_INPUT; + /* UNBLOCK_INPUT; */ if (val) xsignal1 (Qinvalid_regexp, build_string (val)); -- 2.39.2