- Nov 19, 2002
- Nov 18, 2002
-
-
toy authored
Under Linux, the SOFTWARE-VERSION function now returns information on the current kernel taken from /proc/version (instead of "n/a"). Under Linux and xBSD, the internal GET-PAGE-SIZE function obtains the page size from the operating system via the getpagesize() library call, instead of returning a hard-coded value. A few random typos were also fixed.
-
- Nov 15, 2002
-
-
toy authored
Replace the code that grovels /etc/passwd files with an interface to the standard user and group database access functions. This makes CMUCL work better on systems where user databases are accessed using NIS or LDAP. Add functions UNIX-GETPWUID, UNIX-GETPWNAM that return structures of type USER-INFO, and UNIX-GETGRGID and UNIX-GETGRNAM that return structures of type GROUP-INFO. The functions return NIL if the requested information is not available. On Linux/glibc and Solaris, reentrant versions of the functions are used. On FreeBSD we call the non-thread-safe versions.
-
toy authored
-
toy authored
o Adjust rule to make the right index.
-
- Nov 14, 2002
-
-
toy authored
-
toy authored
Noted by Fabricio Chalub.
-
toy authored
The function GENERATE-BYTE-CODE-FOR-REF is calling COMBINATION-ARGS on the return value of CONTINUATION-DEST. In the situation that I have run into, the return value is of type MV-COMBINATION instead of COMBINATION, so an error occurs. Changing the COMBINATION-ARGS to BASIC-COMBINATION-ARGS seems to fix the problem...
-
toy authored
The attached patch contains mostly trivial changes, including a few spelling corrections, plus changes to the types of condition raised in various cases. The only significant change, I believe, is - MAKE-DISPATCH-MACRO-CHARACTER no longer raises an error if the character already was defined to be a macro character; it overrides the previous definition (as per a suggestion from Drew McDermott on cmucl-help, dated 2002-09-22).
-
toy authored
-
toy authored
:ABSOLUTE instead of NIL, as required by CLHS. (I think.)
-
toy authored
o (loop for nil being the hash-keys of (make-hash-table) count t) o (loop for k being the hash-keys of (make-hash-table) using (hash-value (v1 . v2)) count t) From Gerd Moellmann.
-
toy authored
-
- Nov 13, 2002
-
-
toy authored
Bug noted by Paul Dietz, fix by Gerd Moellmann.
-
toy authored
Noted by Brian Spilsbury.
-
toy authored
understand!) SC_OFFSETs in assembly files.
-
toy authored
o Need to restore all global (%g) Lisp registers because they may have been trashed by the C code. o The unimp trap_Error was wrong. Get the right number of arg bytes and the right SC_OFFSET!
-
toy authored
whatever dlsym returns.
-
toy authored
args in the wrong order (stream arg is first). Noted by Matthew Danish.
-
- Nov 12, 2002
-
-
toy authored
not an SAP.
-
- Nov 08, 2002
- Nov 07, 2002
-
-
toy authored
platforms.
-
- Nov 05, 2002
-
-
cracauer authored
Make (time ...) and the profiler do precise measuring of space allocation. It will also not overflow or bomb out when consing amounts cross most-positive fixnum. The new profiler also has an interface to plug in your own print function (also dictates sorting or results). This is written on gencgc/x86 but tests indicated the fallsbacks for other platforms work. The dfixnum package included here is sketchy.
-
- Nov 02, 2002
- Nov 01, 2002
-
-
toy authored
-
toy authored
hold the consing values (to reduce consing in the profiling routines themselves). Based on the double-fixnum idea of Martin Cracauer.
-
toy authored
pushnew screwed up the evaluation order. Fix that, and get the evaluation order correct for multiple-values too.
-
toy authored
-
toy authored
-
toy authored
o Add note about other (unnamed) contributors.
-
toy authored
:propagate-fun-type since that's the default now. o Correct some errors in the examples and text. (Member types and union types are supported.) o Add a section on signed zeroes and special functions. Reference Kahan.
-
- Oct 30, 2002
- Oct 29, 2002
-
-
toy authored
macro or symbol-macro). This allows use of push/pushnew in kernel.core again. Solution suggested by Christophe Rhodes.
-
pmai authored
This patchset by Gerd Moellmann reenables the previous fix to this bug, without retriggering the obsolete-instance-trap-on-redefined-class bug. Additionally, this cleans up the layout-invalid vs. wrapper-state logic making the code much easier to follow.
-
- Oct 28, 2002