Re: Some diffs for MiNT 1.12h3

[mk@anuurn.do.open.de (Martin Koehling)]

Hi Michael,

>Hi!
>
>I'm currently putting together mh-mint-1.12h4.

here are two small fixes for the PLOADINFO handling I forgot to
send to the list...

1. The PLOADINFO code in proc_ioctl() has a one-off bug - it can
   overwrite the end of the user buffer if the load path is exactly
   p->fnamelen bytes long:

--- procfs.c	Thu Nov 24 13:12:14 1994
+++ ..\112h3mk-src\procfs.c	Tue Apr 25 23:35:14 1995
@@ -697,7 +697,7 @@

 			if (!p->fname[0]) return EFILNF;
 			strncpy (pl->cmdlin, p->cmdlin, 128);
-			if (strlen (p->fname) <= pl->fnamelen)
+			if (strlen (p->fname) < pl->fnamelen)
 				strcpy (pl->fname, p->fname);
 			else
 				return ENAMETOOLONG;

2. If a program is started with a relative path *with* a drive letter
   (like "d:foo.ttp"), the drive letter isn't converted to upper case -
   and program code a few lines down relies on it being upper case...

--- dosmem.c	Fri Mar 17 18:56:10 1995
+++ ..\112h3mk-src\dosmem.c	Sat Apr 22 04:05:10 1995
@@ -604,7 +604,7 @@
 			char *source = ptr1;
 			tmp[1] = ':';
 			if (source[1] == ':') {
-				tmp[0] = source[0];
+				tmp[0] = toupper(source[0]);
 				source += 2;
 			} else {
 				tmp[0] = 'A' + curproc->curdrv;

>Has a new Gemini release which fixes that bug already been released?

Not yet; Julian should be able to tell us if the next, soon-to-be-released
(one hopes) version still has this problem...

>I'd prefer not to include this kludge in the release...

Yes, it's a kludge; but it doesn't "cost" all that much - some Pexec
modes are a /tiny bit/ slower, that's all!
(It *would* collide with a possible future Gemini/Mupfel version that
knows about job control; but I don't think something like that will
materialize in the next few months...)

Regards,
     Martin

-- 
 Martin Koehling | mk@anuurn.do.open.de | Martin_Koehling@un.maus.ruhr.de