path component. [1]
- Add the usual MAKE_ENV=LANG=C workaround in order to prevent problems
from bad interactions of gmake, the Schily makefile system and certain
locales.
Requested by: kris [1]
Approved by: netchild
shipping with sformat 3.5 for building sformat; the former include their
own version of __dtoa() which is no longer externally visible with libc
symbol versioning turned on (thus breaking the build of stock sformat).
- While here also use MASTER_SITE_LOCAL for providing a backup mirror of
the sformat 3.5 distfile instead of using a private one.
- Fix the checksumming of sformat.dat on LP64 platforms so sformat doesn't
erroneously complain about that file containing uncertified data there.
Bump PORTREVISION for this.
Approved by: netchild
- Fix display of literal examples in sformat.1; bump PORTREVISION for
this. [1]
- While at the man page remove trailing whitespace, fix spelling errors
and adopt better for FreeBSD.
Reported by: Alex Teclo [1]
Fix by: ru [1]
Approved by: netchild
- Manually replace included copies of config.guess and config.sub with
versions from ${PORTSDIR}/Templates. This is necessary as the sformat
package doesn't have a valid configure script in it's root directory
but one in a sub-directory that is executed by the SING (Schily Is
Not Gnu) makefile system further down the road. Therefor the port
can't set GNU_CONFIGURE which would automatically copy the relevant
files.
- Remove BSD_SCSI_SENSE_BUG from COPTX (CFLAGS) to libscg. It is a
workaround for a (probably alumnus) bug in a SCSI-driver of NetBSD
and the affected code doesn't get compiled on neither pre-CAM nor
CAM versions of FreeBSD.
- Use DATADIR in pkg-plist.
Submitted by: maintainer
bug you are only vulnerable if you locally set the sformat binary suid root.
- Respect CC already at the configure-stage of sformat else we end up running
configure with GCC and compiling with whatever CC was set to.
Submitted by: maintainer
Sun format is a util for formatting/partitioning/analysis/repairing of SCSI
disks. While formatting and partitioning currently is only usefule on big-
endian systems in order to create Sun disk labels, analysis and repairing
works on either endian and are very useful to remap defect blocks when
AWRE/ARRE fail for some reason. Tested on i386 and sparc64, respawned a
disk here.
Submitted by: marius@alchemy.franken.de
