Bruce Evans [Tue, 8 Dec 1998 10:22:07 +0000 (10:22 +0000)]
Backed out the FIOASYNC fix in rev.1.108. fcntl(fd, F_SETFL, flags)
depends on the bug. It does an FIOASYNC ioctl to sync the setting
of the O_ASYNC "file" flag with drivers even if the setting hasn't
changed.
Warner Losh [Mon, 7 Dec 1998 22:36:54 +0000 (22:36 +0000)]
If there is no .depends file, use the standard ad-hoc way of ensuring
that the generated files are generated before any of the object files.
Also minor cleanup of dependencies in conf/files that I bogusly added
before.
This should fix the requirement that make depend be done starting from
a clean config directory. If you don't have a clean directory, make
depend is still required if you want the proper .o's to be recompiled.
Archie Cobbs [Mon, 7 Dec 1998 21:58:50 +0000 (21:58 +0000)]
The "easy" fixes for compiling the kernel -Wunused: remove unreferenced static
and local variables, goto labels, and functions declared but not defined.
Bruce Evans [Mon, 7 Dec 1998 12:14:04 +0000 (12:14 +0000)]
Fixed warnx format errors in printf and csh, and snprintf format errors
in sh, by using separate macros for the 1, 2 and 3-arg calls to warnx.
(The 3-arg warnx macro in sh/bltin/bltin.h used to require bogus dummy
args.)
Bruce Evans [Mon, 7 Dec 1998 10:16:58 +0000 (10:16 +0000)]
Fixed printf format errors (new bugs in rev.1.7). Fixed a spelling error
(rev.1.7 blew away most of rev.1.2-1.6; I'm only fixing blowing away of
rev.1.4).
Bill Paul [Mon, 7 Dec 1998 00:35:06 +0000 (00:35 +0000)]
if_rl.c: make a small tweak to properly handle the Accton MPX 5030/5038
as a RealTek 8139
if_rlreg.h: use bus_space_read_X() in CSR_READ_X() macros instead of
directly calling inb()/outb() etc...
rl.4 + RELNOTES.TXT: mention that SMC EtherEZ PCI 1211-TX is supported
by the RealTek driver
The \*(td string macro, which is supposed to display today's date in the form
"October 14, 1998". A Y2k fix makes it now display "October 14, 98+1900".
Appears you can perform arithmetic in a string definition.
PR: gnu/8321
Reviewed by: Andrew J. Korty <ajk@purdue.edu>
Put mfsroot.gz on yet another filesystem so that the boot blocks
can really read it. Hmm. I have no idea *how* this worked in our
earlier tests on the x86. :)
Matt Jacob [Sat, 5 Dec 1998 23:55:48 +0000 (23:55 +0000)]
Add a CAM_DEBUG_XPT define (to debug XPT layer only). Add a CAM_DEBUGGED
macro to be like the CAM_DEBUG macro but to instead provide a predicate
result.
John Polstra [Sat, 5 Dec 1998 22:15:10 +0000 (22:15 +0000)]
Null commit to commemorate the momentous occasion of the repository
copying of these files from the parent directory. Let us all now
observe a brief moment of silence. OK, that's long enough.
Jordan will adjust the Makefiles appropriately and then "cvs rm" the
originals.
There is a bug in /bin/pax -s option processing. The code assumes that all
pattern matches will occur at offset zero of the source string. The bug causes
the input source string pointer to be incremented by the offset of the end of
the match, instead of it's length. The fix is to only increment the pointer by
the length of the pattern match (eo-so).
Of course, the one example in the man page shows a situation where the match
occurs at offset 0.
Submitted by: John W. DeBoskey <jwd@unx.sas.com>
Obtained from: freebsd-current@freebsd.org
Satoshi Asami [Sat, 5 Dec 1998 06:29:03 +0000 (06:29 +0000)]
Fix support for uncompressed (".tar") package types. It's not completely
fixed (chained dependency checking for pkg_add is broken, for one thing)
but at least you can now create one package and use it.
Kirk McKusick [Sat, 5 Dec 1998 06:12:14 +0000 (06:12 +0000)]
Even the most recently allocated buffer may not have its b_blkno
field properly filled in, so we must do a VOP_BMAP on that buffer
as well if it is not resolved.
Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
Bill Paul [Sat, 5 Dec 1998 02:21:44 +0000 (02:21 +0000)]
Add sanity check to foo_start() routines: in the unlikely (though
apparently possible) event that the transmit start routine is
called with and empty if_snd queue, bail out instead of dereferencing
unilitialized transmit list pointers and panicking.
Matt Jacob [Sat, 5 Dec 1998 01:33:57 +0000 (01:33 +0000)]
Remove the Target mode functions until they're in better shape. Implement some
suggested compilation cleanups from Eklund. Wire down a hard loop id if we are
not on a platform that has the ability to get to a PCI BIOS (it still will
float to the ID it gets after a LIP but at least we can try). Clarify that the
expanded lun is based upon SCCLUN defines (in f/w).
Matt Jacob [Sat, 5 Dec 1998 00:47:25 +0000 (00:47 +0000)]
Roll to 1.15 f/w level for Qlogic 2100. Leave all flavors in, but control which
one gets compiled in by default. The default is private loop, non-expanded lun.
Archie Cobbs [Fri, 4 Dec 1998 22:54:57 +0000 (22:54 +0000)]
Examine all occurrences of sprintf(), strcat(), and str[n]cpy()
for possible buffer overflow problems. Replaced most sprintf()'s
with snprintf(); for others cases, added terminating NUL bytes where
appropriate, replaced constants like "16" with sizeof(), etc.
These changes include several bug fixes, but most changes are for
maintainability's sake. Any instance where it wasn't "immediately
obvious" that a buffer overflow could not occur was made safer.
Reviewed by: Bruce Evans <bde@zeta.org.au>
Reviewed by: Matthew Dillon <dillon@apollo.backplane.com>
Reviewed by: Mike Spengler <mks@networkcs.com>
Bill Paul [Fri, 4 Dec 1998 21:48:12 +0000 (21:48 +0000)]
Put back some missing chip identification code that got lost somehow
during a trek through RCS. The Macronix 98713 and 98713A both have the
same PCI device ID but different revision numbers, and we need to be
able to tell one from the other. The 98715 and 98715A chips have the
same device ID as the 98725 chip but different revision numbers, however
we lump them into the same category except when identifying them during
the PCI probe output.
The main reason we need tell the chips apart is that the Macronix app
notes say you have to write a special magic number into one of the
registers in order to put the chip in normal operating mode. The 98713
requires one magic value, while all the others require a different one.
Archie Cobbs [Fri, 4 Dec 1998 21:28:39 +0000 (21:28 +0000)]
Fix typo: "==" should have been "="
PR: 8280 (1/3 patches contained in this PR)
Reviewed by: Nate Williams <nate@mt.sri.com>
Submitted by: Sakari Jalovaara <sja@tekla.fi>
Bill Paul [Fri, 4 Dec 1998 18:01:24 +0000 (18:01 +0000)]
An early Christmas present: add driver support for a whole bunch of
PCI fast ethernet adapters, plus man pages.
if_pn.c: Netgear FA310TX model D1, LinkSys LNE100TX, Matrox FastNIC 10/100,
various other PNIC devices
if_mx.c: NDC Communications SOHOware SFA100 (Macronix 98713A), various
other boards based on the Macronix 98713, 98713A, 98715, 98715A
and 98725 chips
if_vr.c: D-Link DFE530-TX, other boards based on the VIA Rhine and
Rhine II chips (note: the D-Link and certain other cards
that actually use a Rhine II chip still return the PCI
device ID of the Rhine I. I don't know why, and it doesn't
really matter since the driver treats both chips the same
anyway.)
if_wb.c: Trendware TE100-PCIE and various other cards based on the
Winbond W89C840F chip (the Trendware card is identical to
the sample boards Winbond sent me, so who knows how many
clones there are running around)
All drivers include support for ifmedia, BPF and hardware multicast
filtering.
Also updated GENERIC, LINT, RELNOTES.TXT, userconfig and
sysinstall device list.
I also have a driver for the ASIX AX88140A in the works.
Doug Rabson [Fri, 4 Dec 1998 10:52:48 +0000 (10:52 +0000)]
Implement 'software completion' for floating point arithmetic. On the
alpha, operations involving non-finite numbers or denormalised numbers
or operations which should generate such numbers will cause an arithmetic
exception. For programs which follow some strict code generation rules,
the kernel trap handler can then 'complete' the operation by emulating
the faulting instruction.
To use software completion, a program must be compiled with the arguments
'-mtrap-precision=i' and '-mfp-trap-mode=su' or '-mfp-trap-mode=sui'.
Programs compiled in this way can use non-finite and denormalised numbers
at the expense of slightly less efficient code generation of floating
point instructions. Programs not compiled with these options will receive
a SIGFPE signal when non-finite or denormalised numbers are used or
generated.
John Birrell [Thu, 3 Dec 1998 23:02:03 +0000 (23:02 +0000)]
Add __attribute__ ((unused)) to the SYSINIT etc macros which declare
static structures that are used with the data set magic. This allows
kernel modules, for example, to be compiled with -Wall -Werror.