Charles Bailey hinted at alignment problems with the piping code
on his VAX; I think I got it tracked down now...it seems that some
stuff in one of my structures has a different size on VAX and Alpha.
Moving the alignment-sensitive queue stuff to the beginning of the
structure seems to do the trick.
In the test suite, this manifests itself as an error in
the io/openpid test #3.
There's also a trivial problem with prototyping LIB$ and SYS$
functions on the VAX: it wants lowercase only; I'm leaving that
out of the patch for now.
This patch is to be applied after the piping fix of 3/28, for
those who can't wait...otherwise I'm going to try combine the
three piping fixes into a single jumbo patch and send it out
in the next few days.
--- vms/vms.c-orig Tue May 9 15:40:56 2000
+++ vms/vms.c Tue May 9 09:53:46 2000
@@ -1037,6 +1037,8 @@
};
struct _pipe {
+ RQE free;
+ RQE wait;
int fd_out;
unsigned short chan_in;
unsigned short chan_out;
@@ -1053,8 +1055,6 @@
pInfo info;
pCBuf curr;
pCBuf curr2;
- RQE free;
- RQE wait;
};
--
Drexel University \V --Chuck Lane
----------------->--------*------------<[EMAIL PROTECTED]
(215) 895-1545 / \ Particle Physics [EMAIL PROTECTED]
FAX: (215) 895-5934 /~~~~~~~~~~~ [EMAIL PROTECTED]
