d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
148,304 Commits 6 Branches 133 Tags
Commit Graph

73283 Commits

Author SHA1 Message Date
Oleksandr Tymoshenko
6846a68073 - Mark temp variable as "earlyclobber" in assembler inline in 
atomic_fetchadd_32.  Without it gcc would use it as input
    register for v and sometimes generate following code for
    function call like atomic_fetchadd_32(&(fp)->f_count, -1):

801238b4:       2402ffff        li      v0,-1
801238b8:       c2230018        ll      v1,24(s1)
801238bc:       00431021        addu    v0,v0,v1
801238c0:       e2220018        sc      v0,24(s1)
801238c4:       1040fffc        beqz    v0,801238b8 <dupfdopen+0x2e8>
801238c8:       00000000        nop

   Which is definitly wrong because if sc fails v0 is set to 0
   and previous value of -1 is overriden hence whole operation
   turns to bogus
 2009-06-19 04:43:49 +00:00
Oleksandr Tymoshenko
1a28ce2dcb - Remove debug printfs 2009-06-16 02:36:21 +00:00
Oleksandr Tymoshenko
04af8c86f5 - Handle KSEG0/KSEG1 addresses for /dev/mem as well. netstat requires it 2009-06-16 01:43:33 +00:00
Oleksandr Tymoshenko
8766ab738d - Take into account only unmasked bits in interrupt status register 2009-06-16 00:02:02 +00:00
Oleksandr Tymoshenko
ee58bf3153 - Fix prototypes to make compiler happy 2009-06-14 21:16:23 +00:00
Oleksandr Tymoshenko
d7913bf5ab - Get rid of mask_fn and fix pre_filter/post_filter functions' prototypes 2009-06-14 21:16:04 +00:00
Oleksandr Tymoshenko
1ee48ffda9 - Fix prototype and implementation of admsw_shutdown 2009-06-14 21:04:54 +00:00
Oleksandr Tymoshenko
e3652db0ce - Fix prototypes to make compiler happy 2009-06-14 20:54:46 +00:00
Warner Losh
72bd8c62f6 Minor formatting changes. Also, elimiante a couple of unused variables. 2009-06-14 07:01:22 +00:00
Warner Losh
c839424d34 Various nits to make this compile. 2009-06-14 06:53:55 +00:00
Warner Losh
24277e95bb Make compile. 2009-06-14 06:49:13 +00:00
Warner Losh
67b401589c Hack for the 'battleship' boards that have 8 ports rather than 4. 2009-06-14 06:36:50 +00:00
Warner Losh
bafe55344b Compile out unreferenced code. 2009-06-14 06:35:21 +00:00
Warner Losh
db9c08f280 Stylish nits 2009-06-14 06:35:02 +00:00
Warner Losh
00e1958bd4 Add bogus OCTEON_CORE_ID here. Really should integrate the pcpu.h stuff 
that is in Cavium's base port.
 2009-06-14 06:27:11 +00:00
Warner Losh
c453afd113 this was actually deleted earlier, but I bogusly neglected to commit it. 2009-06-14 06:26:33 +00:00
Warner Losh
2e24c40fec bye bye. This is no longer referenced, but much code from it will 
resurface for a bus-space implementation.
 2009-06-14 06:14:25 +00:00
Warner Losh
78e0ef6f04 Kludge: pretend to be ISA_MIPS32 for the moment. 2009-06-14 06:12:21 +00:00
Warner Losh
2acee4de5c First pass to make compile. It doesn't completely yet, but it's a 
start.
 2009-06-14 06:11:51 +00:00
Warner Losh
8c29759b4a Bring back the TARGET_OCTEON kludge for a bit. We need to kill it, 
but it is useful for the moment.
 2009-06-14 06:11:13 +00:00
Warner Losh
55f888f68d Cavium-specific goo is no longer necessary here. Of course, I now 
have to write a bus space for cavium, but that shouldn't be too hard.
 2009-06-14 06:10:36 +00:00
Warner Losh
29854186d2 Move octeon specific uart goo here, per SOP for other MIPS ports. 2009-06-14 06:09:33 +00:00
Warner Losh
558955d6a8 Move this to a more approrpiate plae. 2009-06-14 06:01:46 +00:00
Warner Losh
f7edf22465 Bring this in from the cavium port. 2009-06-14 05:29:13 +00:00
Warner Losh
0d633f654d o Move the driveid.h file 
o lots of tweaks to header paths.
o comment out SMP for the moment

# we now make it through the .c make depend, the .s needs more work.
 2009-06-14 04:26:56 +00:00
Warner Losh
d0b7b80503 Part of the cf driver missed. 2009-06-14 04:16:58 +00:00
Warner Losh
654d4c2496 Move octeon rgmii driver to is more correct new home. 2009-06-14 04:10:27 +00:00
Warner Losh
5a2edc8c75 Import sgmii driver for Octeon from Cavium FreeBSD tree. 2009-06-14 03:57:03 +00:00
Warner Losh
b02713c7a4 Move dev/flash/ cf driver into octeon dir where it belongs. 2009-06-14 03:55:27 +00:00
Warner Losh
5be117dfac Merge in Cavium's CF driver. This too is in the wrong place and will 
be moved.
 2009-06-14 03:47:44 +00:00
Warner Losh
b9bf0e01e9 Actually rename the files this time. Also, start to fix OCTEON1 so it 
can configure.
 2009-06-14 03:44:43 +00:00
Warner Losh
6aca1d2ff0 Now that the import is over, we can (re) delete this. 2009-06-14 03:44:01 +00:00
Warner Losh
b502e57d9e Move the octeon port to its more correct location. Any port for the 
OCTEON2 family of processors should live in mips/octeon2.  Not enough
is know abotu the former to know if the same port can be used for both
yet.
 2009-06-14 03:01:39 +00:00
Warner Losh
75c1942047 Import Cavium's FreeBSD port, or the Octeon specific pieces, verbatim. 
Yes, this puts things in the wrong place, doesn't compile and is
woefully incomplete.  However, it will allow us to more easily track
against the upstream sources without needing to import the entire
Cavium tree under vendor.

This port is based on FreeBSD 7.0 as of April 2007 and the pre-import
MIPS tree (aka mips2), so much work is necessary here.
 2009-06-14 02:46:07 +00:00
Oleksandr Tymoshenko
fdf7155886 - Fix functions prototypes to make compiler happy 2009-06-12 12:17:32 +00:00
Oleksandr Tymoshenko
b64bb4d93e - Switch no normal RPC oimplementation. Unaligned access bug has 
been fixed
 2009-06-12 12:10:10 +00:00
Oleksandr Tymoshenko
7e857dd14f - Merge from HEAD 2009-06-12 11:50:44 +00:00
Marko Zec
878a6d7dff Remove unnecessary #ifdef lines and code. 
Approved by:	julian (mentor)
 2009-06-12 09:31:14 +00:00
Marko Zec
eb4261851f Assign ng_eiface nodes a netgraph name on instantiation, in a way which 
is consistent with the current behavior of ng_iface, i.e. borrow the
same naming code from ng_iface.c.

Approved by:	julian (mentor)
 2009-06-12 09:20:31 +00:00
Andriy Gapon
925b1b25f1 KOBJMETHOD: throw the switch on kobj method signature checking 
It seems that current code should pass the check.
This commit should not lead to any changes in compiled code.
From now on a warning shall be produced if kobj method implementation
function has a mismatching signature.

Verified by:	md5
Reviewed by:	imp
Approved by:	jhb (mentor)
 2009-06-12 09:05:23 +00:00
Ariff Abdullah
f6f8d7b601 Last (hopefully) attempt to fix both parallel build and 
cyclic dependencies, in a simmilar fashion to kmod.mk.

Discussed with:		bde, bz
 2009-06-12 07:39:10 +00:00
Warner Losh
b27c946750 devclass_find_free_unit(xxx, 0) is identical to -1 for most 
applications (including this one).  Switch to it since the former is a
racy interface.
 2009-06-12 00:07:09 +00:00
Kip Macy
f0c6b798a3 pjd has requested that I keep the tunable as zfs_prefetch_disable to minimize gratuitous 
differences with Opensolaris' ZFS

Sorry for the churn
 2009-06-11 22:24:08 +00:00
George V. Neville-Neil
02c7d9a64f Re-add the send queue tunable for people who do not use buffering. 
Reviewed by:	jhb
MFC after:	3 days
 2009-06-11 21:32:26 +00:00
Andriy Gapon
198f4cc7b6 KOBJMETHOD: use better construct in the checking version of the macro 
Big thanks to Christoph Mallon for the idea/code!
This construct has benefit of sticking much stricter to C standard and thus
keeping more compilers happy as Clang doesn't like the current construct
because it doesn't treat FUNC != NULL as a compile-time constant.

The checking version is still under 'notyet'.

Pointed out by:	ed
Submitted by:	Christoph Mallon <christoph.mallon@gmx.de>
Clang help by:	rdivacky
Reviewed by:	imp
Approved by:	jhb
 2009-06-11 17:21:14 +00:00
Andriy Gapon
65f4467938 strict kobj signatures: fix adb_hb_controller_poll impl in powermac 
the method return u_int, not void

Reviewed by:	imp, current@
Approved by:	jhb (mentor)
 2009-06-11 17:16:04 +00:00
Andriy Gapon
1c44eb75f7 strict kobj signatures: fix iicbus_write impl in dev/pcf 
input buf is const

Reviewed by:	imp, current@
Approved by:	jhb (mentor)
 2009-06-11 17:15:44 +00:00
Andriy Gapon
4fc23012c6 strict kobj signatures: some ofw_setprop fixes 
propname parameter is const

Reviewed by:	imp, current@
Approved by:	jhb (mentor)
 2009-06-11 17:15:20 +00:00
Andriy Gapon
8e45f0b7c6 strict kobj signatures: fix assortment of miibus_writereg impls 
return type should be int, not void

Reviewed by:	imp, current@
Approved by:	jhb (mentor)
 2009-06-11 17:14:54 +00:00
Andriy Gapon
c0e55e3b65 strict kobj sigs: fix assortment of device_detach and device_shutdown impls 
with common issue of having void return type instead of int

Reviewed by:	imp, current@
Approved by:	jhb (mentor)
 2009-06-11 17:14:28 +00:00