Zsh Mailing List Archive
Messages sorted by: Reverse Date, Date, Thread, Author

Re: crash/hang with gcc 5+ -O2 and --enable-zsh-mem



On Sat, Feb 24, 2018 at 02:13:34PM -1000, Joey Pabalinas wrote:
> The only fix I could find which didn't requiring substantial
> reimplementation of the memory management functions was to replace
> the malloc() call in calloc() with realloc() instead. With a NULL `p`
> argument realloc() behaves exactly the same as malloc() does, and
> (at least on my system) gcc doesn't seem to consider realloc() a
> candidate for sibling call optimizations; give this patch a try
> and _hopefully_ this is a viable solution.

On second thought, doing it this way is probably a *little* bit better; the
needless initialization of `r` to NULL is avoided, and it also makes the
purpose of using realloc() over malloc() a *tiny* bit more explicit:

Signed-off-by: Joey Pabalinas <joeypabalinas@xxxxxxxxx>

 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/Src/mem.c b/Src/mem.c
index 840bbb6e4a4eb6fd73..f1208197b3ddac2139 100644
--- a/Src/mem.c
+++ b/Src/mem.c
@@ -1719,7 +1719,13 @@ calloc(MALLOC_ARG_T n, MALLOC_ARG_T size)
     if (!(l = n * size))
 	return (MALLOC_RET_T) m_high;
 
-    r = malloc(l);
+    /*
+     * use realloc() (with a NULL `p` argument it behaves exactly the same
+     * as malloc() does) to prevent an infinite loop caused by sibling-call
+     * optimizations (the malloc() call would otherwise be replaced by an
+     * unconditional branch back to line 1719 ad infinitum).
+     */
+    r = realloc(NULL, l);
 
     memset(r, 0, l);
 
-- 
2.16.2

Attachment: signature.asc
Description: PGP signature



Messages sorted by: Reverse Date, Date, Thread, Author