[2/3] libc: Remove support for pre-C99 C standards

Message ID 20260620053946.556799-3-sebastian.huber@embedded-brains.de
State New
Headers
Series Synchronize div() implementation with FreeBSD |

Commit Message

Sebastian Huber June 20, 2026, 5:39 a.m. UTC
  From: Minsoo Choo <minsoochoo0122@proton.me>

Reviewed by:	jhb
Differential Revision:	https://reviews.freebsd.org/D43254
---
 newlib/libc/stdlib/div.c | 35 +----------------------------------
 1 file changed, 1 insertion(+), 34 deletions(-)
  

Patch

diff --git a/newlib/libc/stdlib/div.c b/newlib/libc/stdlib/div.c
index 4296f1e80..467f25aa3 100644
--- a/newlib/libc/stdlib/div.c
+++ b/newlib/libc/stdlib/div.c
@@ -86,39 +86,6 @@  div (int num,
 
 	r.quot = num / denom;
 	r.rem = num % denom;
-	/*
-	 * The ANSI standard says that |r.quot| <= |n/d|, where
-	 * n/d is to be computed in infinite precision.  In other
-	 * words, we should always truncate the quotient towards
-	 * 0, never -infinity or +infinity.
-	 *
-	 * Machine division and remainer may work either way when
-	 * one or both of n or d is negative.  If only one is
-	 * negative and r.quot has been truncated towards -inf,
-	 * r.rem will have the same sign as denom and the opposite
-	 * sign of num; if both are negative and r.quot has been
-	 * truncated towards -inf, r.rem will be positive (will
-	 * have the opposite sign of num).  These are considered
-	 * `wrong'.
-	 *
-	 * If both are num and denom are positive, r will always
-	 * be positive.
-	 *
-	 * This all boils down to:
-	 *	if num >= 0, but r.rem < 0, we got the wrong answer.
-	 * In that case, to get the right answer, add 1 to r.quot and
-	 * subtract denom from r.rem.
-	 *      if num < 0, but r.rem > 0, we also have the wrong answer.
-	 * In this case, to get the right answer, subtract 1 from r.quot and
-	 * add denom to r.rem.
-	 */
-	if (num >= 0 && r.rem < 0) {
-		++r.quot;
-		r.rem -= denom;
-	}
-	else if (num < 0 && r.rem > 0) {
-		--r.quot;
-		r.rem += denom;
-	}
+
 	return (r);
 }