From mboxrd@z Thu Jan 1 00:00:00 1970 X-Msuck: nntp://news.gmane.org/gmane.linux.lib.musl.general/13296 Path: news.gmane.org!.POSTED!not-for-mail From: Rich Felker Newsgroups: gmane.linux.lib.musl.general Subject: un-UBify-strings Date: Sat, 22 Sep 2018 20:35:42 -0400 Message-ID: <20180923003542.GC17995@brightrain.aerifal.cx> Reply-To: musl@lists.openwall.com NNTP-Posting-Host: blaine.gmane.org Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="lteA1dqeVaWQ9QQl" X-Trace: blaine.gmane.org 1537662832 8029 195.159.176.226 (23 Sep 2018 00:33:52 GMT) X-Complaints-To: usenet@blaine.gmane.org NNTP-Posting-Date: Sun, 23 Sep 2018 00:33:52 +0000 (UTC) User-Agent: Mutt/1.5.21 (2010-09-15) To: musl@lists.openwall.com Original-X-From: musl-return-13312-gllmg-musl=m.gmane.org@lists.openwall.com Sun Sep 23 02:33:48 2018 Return-path: Envelope-to: gllmg-musl@m.gmane.org Original-Received: from mother.openwall.net ([195.42.179.200]) by blaine.gmane.org with smtp (Exim 4.84_2) (envelope-from ) id 1g3sLL-00020Q-Rn for gllmg-musl@m.gmane.org; Sun, 23 Sep 2018 02:33:47 +0200 Original-Received: (qmail 19560 invoked by uid 550); 23 Sep 2018 00:35:55 -0000 Mailing-List: contact musl-help@lists.openwall.com; run by ezmlm Precedence: bulk List-Post: List-Help: List-Unsubscribe: List-Subscribe: List-ID: Original-Received: (qmail 19525 invoked from network); 23 Sep 2018 00:35:55 -0000 Content-Disposition: inline Original-Sender: Rich Felker Xref: news.gmane.org gmane.linux.lib.musl.general:13296 Archived-At: --lteA1dqeVaWQ9QQl Content-Type: text/plain; charset=us-ascii Content-Disposition: inline I've had this patch sitting around since 2016, and just updated it to apply cleanly. Any objections? Since I killed the stdio UB in this release cycle I'd like to go ahead and eliminate all the string-function UB that can be eliminated (there's still aligned read past end of string that's unfixable without an attribute that explicitly allows it, or asm; it might turn out that asm would make sense here). Rich --lteA1dqeVaWQ9QQl Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename="un-UB-strings.diff" diff --git a/src/string/memccpy.c b/src/string/memccpy.c index 7c233d5..5c8b672 100644 --- a/src/string/memccpy.c +++ b/src/string/memccpy.c @@ -11,19 +11,21 @@ void *memccpy(void *restrict dest, const void *restrict src, int c, size_t n) { unsigned char *d = dest; const unsigned char *s = src; - size_t *wd, k; - const size_t *ws; c = (unsigned char)c; +#ifdef __GNUC__ + size_t __attribute__((__may_alias__)) *wd; + const size_t __attribute__((__may_alias__)) *ws; if (((uintptr_t)s & ALIGN) == ((uintptr_t)d & ALIGN)) { for (; ((uintptr_t)s & ALIGN) && n && (*d=*s)!=c; n--, s++, d++); if ((uintptr_t)s & ALIGN) goto tail; - k = ONES * c; + size_t k = ONES * c; wd=(void *)d; ws=(const void *)s; for (; n>=sizeof(size_t) && !HASZERO(*ws^k); n-=sizeof(size_t), ws++, wd++) *wd = *ws; d=(void *)wd; s=(const void *)ws; } +#endif for (; n && (*d=*s)!=c; n--, s++, d++); tail: if (*s==c) return d+1; diff --git a/src/string/memchr.c b/src/string/memchr.c index 4daff7b..1038ce6 100644 --- a/src/string/memchr.c +++ b/src/string/memchr.c @@ -12,12 +12,14 @@ void *memchr(const void *src, int c, size_t n) { const unsigned char *s = src; c = (unsigned char)c; + +#ifdef __GNUC__ for (; ((uintptr_t)s & ALIGN) && n && *s != c; s++, n--); - if (n && *s != c) { - const size_t *w; - size_t k = ONES * c; - for (w = (const void *)s; n>=SS && !HASZERO(*w^k); w++, n-=SS); - for (s = (const void *)w; n && *s != c; s++, n--); - } + const __attribute__((__may_alias__)) size_t *w; + size_t k = ONES * c; + for (w = (const void *)s; n>=SS && !HASZERO(*w^k); w++, n-=SS); + s = (const void *)w; +#endif + for (; n && *s != c; s++, n--); return n ? (void *)s : 0; } diff --git a/src/string/stpcpy.c b/src/string/stpcpy.c index 54cf9ca..f115d16 100644 --- a/src/string/stpcpy.c +++ b/src/string/stpcpy.c @@ -9,9 +9,9 @@ char *__stpcpy(char *restrict d, const char *restrict s) { - size_t *wd; - const size_t *ws; - +#ifdef __GNUC__ + size_t __attribute__((__may_alias__)) *wd; + const size_t __attribute__((__may_alias__)) *ws; if ((uintptr_t)s % ALIGN == (uintptr_t)d % ALIGN) { for (; (uintptr_t)s % ALIGN; s++, d++) if (!(*d=*s)) return d; @@ -19,6 +19,7 @@ char *__stpcpy(char *restrict d, const char *restrict s) for (; !HASZERO(*ws); *wd++ = *ws++); d=(void *)wd; s=(const void *)ws; } +#endif for (; (*d=*s); s++, d++); return d; diff --git a/src/string/stpncpy.c b/src/string/stpncpy.c index d6d92ff..099d77c 100644 --- a/src/string/stpncpy.c +++ b/src/string/stpncpy.c @@ -9,9 +9,9 @@ char *__stpncpy(char *restrict d, const char *restrict s, size_t n) { - size_t *wd; - const size_t *ws; - +#ifdef __GNUC__ + size_t __attribute__((__may_alias__)) *wd; + const size_t __attribute__((__may_alias__)) *ws; if (((uintptr_t)s & ALIGN) == ((uintptr_t)d & ALIGN)) { for (; ((uintptr_t)s & ALIGN) && n && (*d=*s); n--, s++, d++); if (!n || !*s) goto tail; @@ -20,6 +20,7 @@ char *__stpncpy(char *restrict d, const char *restrict s, size_t n) n-=sizeof(size_t), ws++, wd++) *wd = *ws; d=(void *)wd; s=(const void *)ws; } +#endif for (; n && (*d=*s); n--, s++, d++); tail: memset(d, 0, n); diff --git a/src/string/strchrnul.c b/src/string/strchrnul.c index f2b9ae1..6875ae0 100644 --- a/src/string/strchrnul.c +++ b/src/string/strchrnul.c @@ -9,16 +9,18 @@ char *__strchrnul(const char *s, int c) { - size_t *w, k; - c = (unsigned char)c; if (!c) return (char *)s + strlen(s); +#ifdef __GNUC__ + size_t __attribute__((__may_alias__)) *w; for (; (uintptr_t)s % ALIGN; s++) if (!*s || *(unsigned char *)s == c) return (char *)s; - k = ONES * c; + size_t k = ONES * c; for (w = (void *)s; !HASZERO(*w) && !HASZERO(*w^k); w++); - for (s = (void *)w; *s && *(unsigned char *)s != c; s++); + s = (void *)w; +#endif + for (; *s && *(unsigned char *)s != c; s++); return (char *)s; } diff --git a/src/string/strlcpy.c b/src/string/strlcpy.c index dcb22f6..a76b7b2 100644 --- a/src/string/strlcpy.c +++ b/src/string/strlcpy.c @@ -12,9 +12,10 @@ size_t strlcpy(char *d, const char *s, size_t n) { char *d0 = d; size_t *wd; - const size_t *ws; if (!n--) goto finish; +#ifdef __GNUC__ + const __attribute__((__may_alias__)) size_t *ws; if (((uintptr_t)s & ALIGN) == ((uintptr_t)d & ALIGN)) { for (; ((uintptr_t)s & ALIGN) && n && (*d=*s); n--, s++, d++); if (n && *s) { @@ -24,6 +25,7 @@ size_t strlcpy(char *d, const char *s, size_t n) d=(void *)wd; s=(const void *)ws; } } +#endif for (; n && (*d=*s); n--, s++, d++); *d = 0; finish: diff --git a/src/string/strlen.c b/src/string/strlen.c index 929ddcb..27b6d37 100644 --- a/src/string/strlen.c +++ b/src/string/strlen.c @@ -10,9 +10,12 @@ size_t strlen(const char *s) { const char *a = s; - const size_t *w; +#ifdef __GNUC__ + const __attribute__((__may_alias__)) size_t *w; for (; (uintptr_t)s % ALIGN; s++) if (!*s) return s-a; for (w = (const void *)s; !HASZERO(*w); w++); - for (s = (const void *)w; *s; s++); + s = (const void *)w; +#endif + for (; *s; s++); return s-a; } --lteA1dqeVaWQ9QQl--