Re: remaining signedness issue in Encode 1.88

On Thursday, Feb 27, 2003, at 16:56 Asia/Tokyo, Dan Kogai wrote:
> On Wednesday, Feb 26, 2003, at 14:49 Asia/Tokyo, Craig A. Berry wrote:
>> it seems like it would be a good idea to protect against that even if
>> it's not supposed to happen.
>
> I do agree with that.  The question is how...

Craig,

   Would you try the patch below ?  Instead of typecasting I now made 
sure that slen is never negative.  Already passed 'make test' here.

--- Encode.xs   2003/02/20 14:42:34     1.54
+++ Encode.xs   2003/02/27 09:32:04
@@ -78,11 +78,15 @@

      if (offset) {
        s += *offset;
-      slen -= *offset;
+      if (slen > *offset){ /* safegurd against slen overflow */
+         slen -= *offset;
+      }else{
+         slen = 0;
+      }
        tlen = slen;
      }

-    if (slen <= 0){
+    if (slen == 0){
         SvCUR_set(dst, 0);
         SvPOK_only(dst);
         goto ENCODE_END;
__END_OF_PATCH__

(U8 *)"Dan the Encode Maintainer";

• remaining signedness issue in Encode 1.88 by Craig A. Berry
• Re: remaining signedness issue in Encode 1.88 by Dan Kogai
• Re: remaining signedness issue in Encode 1.88 by Dan Kogai
• Re: remaining signedness issue in Encode 1.88 by Craig A. Berry
• [Encode] 1.89 Released by Dan Kogai
• Re: remaining signedness issue in Encode 1.88 by Dan Kogai