[Bug 548607] Review Request: pvs-sbcl - SRI's Prototype Verification System

[bugzilla at redhat.com]

Please do not reply directly to this email. All additional
comments should be made in the comments box of this bug.


https://bugzilla.redhat.com/show_bug.cgi?id=548607





--- Comment #9 from Jerry James <loganjerry at gmail.com>  2009-12-22 12:44:12 EDT ---
Comment 5: I did tell Sam Owre that I was doing this.  The comment in the spec
file about not sending it upstream is therefore not completely accurate; I just
don't expect upstream to apply it.  They are concentrated on providing a single
download for researchers to grab, unpack, and get to work.  Finding
dependencies would just complicate matters for them.

I renamed the binary from pvs to pvs-sbcl to avoid a name collision (see my
response to comment 2 in comment 4).  So when I talked about the Provide, I was
really only talking about the Provide, not the binary name.

The build hang has nothing to do with 32-bit vs. 64-bit systems, although I
have never seen that hang with 64-bit Emacs (but have with 64-bit XEmacs).  The
problem is that (X)Emacs is forking off a subprocess to run PVS, but isn't
noticing when that process exits, and therefore waits forever.  Stephen
Turnbull believes that this is a bug in X (see the URL in comment 4).  In any
case, I believe I have found a workaround to the problem.  Briefly, I'm
replacing (sit-for 1), which doesn't always notice changes in process status,
with (accept-process-output nil 1), which does.

Comments 6 through 8: I have applied the "|| true" change to the spec file to
avoid killing the build with an unprelinked sbcl binary.

New URLs:
http://jjames.fedorapeople.org/pvs/pvs-sbcl.spec
http://jjames.fedorapeople.org/pvs/pvs-sbcl-4.2-2.20091008svn.fc12.src.rpm

-- 
Configure bugmail: https://bugzilla.redhat.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.