Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
166,299 Commits 606 Branches 0 Tags 3.1 GiB
Commit Graph

166299 Commits

Author SHA1 Message Date
dholland 398690b6f4 Call srandom() as well as initializing the private random generator 
(which is crappy and should be removed, but that's for another day...)
This way e.g. fortune cookies no longer appear always in the same order.
 2008-01-28 05:48:57 +00:00
dholland 9cc7e27481 Larn now builds with WARNS=4. 2008-01-28 05:38:53 +00:00
dholland 3d81345939 Apparently the reason the printf arguments were all wrong is that larn 
had its own private and not-quite-standard implementation of printf.
Remove it surgically with a chainsaw.
 2008-01-28 04:04:17 +00:00
dholland 8ce4d17196 Add gcc printf format checking, and fix the abundant problems this revealed. 
(It appears that someone sometime thought that you use %d to print a long.)
 2008-01-28 03:39:30 +00:00
christos 20275a81fb mention posix_fadvise lossage. 2008-01-28 03:30:26 +00:00
dholland 733dc966d8 clean for WARNS=4 2008-01-28 03:23:29 +00:00
rumble ff2d046b84 Fix MD set tweaking by introducing `md_init_set_status', which is called 
at the end of `init_set_status'.

Previously, i386/amd64, mac68k and sgimips were using `md_init' to set the
appropriate kernel, but any changes made were being undone by a later call
to `init_set_status'.

This should fix IP3x installation issues on sgimips.

XXX - `md_init' is now unused. Should we G/C it?
 2008-01-28 02:47:12 +00:00
dholland 3bf4735ba0 Work around gcc signedness warning: if you check for uid_t < 0, gcc warns 
because it's unsigned, but I don't really want to rely on uid_t being
unsigned on every platform this code might propagate to.
 2008-01-28 02:37:50 +00:00
dholland a627b0395f Build with WARNS=4. Sort out the game's shadowing of its own variables, 
which fixes at least one bug.
 2008-01-28 01:58:01 +00:00
jmcneill 05b25a6836 Ensure that we don't clear the EHCI_PS_CLEAR bits by writing 1s to them 
when modifying port state in ehci_suspend and ehci_resume.
 2008-01-28 01:52:31 +00:00
dholland bded8baaa8 Build with WARNS=4. (All fixes are cosmetic, except that it's definitely 
bad karma to shadow a local variable with another of the same name in a
nested block...)
 2008-01-28 01:38:59 +00:00
dholland 80def703dc Build with WARNS=4. 2008-01-28 01:15:55 +00:00
dholland 68bfa787b0 This builds with WARNS=4; set it in the Makefile so it stays that way. 2008-01-28 01:10:23 +00:00
christos ef1144e6de Always try both tcp and udp, trying first the one specified by the protocol 
wanted.
 2008-01-28 01:10:22 +00:00
dholland d463e4dba6 Build with WARNS=4. 2008-01-28 01:06:19 +00:00
jmcneill e4e10d4f97 Release port ownership on detach, so devices will be routed to companion 
controllers. This allows 'drvctl -d ehci0' to disable USB 2.0 mode.
 2008-01-28 00:47:05 +00:00
jmcneill 515d26c53b Release ownership of EHCI to the BIOS when we suspend or detach the 
device driver.
 2008-01-28 00:44:17 +00:00
martin e68297ad60 wrong argument type 2008-01-28 00:19:28 +00:00
pooka eb6074e41f Replace vrelel() 010101-mania with a flags parameter. However, 
leave flags unimplemented for a while (no change in functionality).
 2008-01-27 22:47:31 +00:00
ad b9839de884 Replace use of lockmgr(). 2008-01-27 21:58:39 +00:00
christos dedb237183 Remove static module defines; they are done in openpam.h now. 2008-01-27 20:26:56 +00:00
martin 98e544d2b4 Regen (posix_fadvise versionend) 2008-01-27 20:14:12 +00:00
martin c9146bcc25 Catch up with versioning of posix_fadvise 2008-01-27 20:13:07 +00:00
pooka 2ff7c33801 honor pager locking protocol in uao_detach() 2008-01-27 20:10:53 +00:00
jmcneill 31646d87c2 Don't attempt to claim ownership from the BIOS if the OS already holds 
the HC OS Owned semaphore. Unconditionally clear SMIs as long as the
USB legacy capability is present.
 2008-01-27 20:09:14 +00:00
jmcneill b6294fbf32 Add some USB EHCI Legacy Support Extended CSR bit definitions. 2008-01-27 20:06:48 +00:00
pooka 18e91c156a Don't fool around when recycling vnodes, simply override refcount 
to 1, hold on tight to your hat, and call recycle.  The reason is
that we might pick up extra references such as from lookup and we
don't really care about tracking those.  For example with p2k,
puffs in the kernel already does all the refcounting we could ever
dream of, and therefore it's left to the client and unnecessary.
 2008-01-27 20:01:29 +00:00
dsl cb0033a159 Call do_posix_fadvise() in netbsd32_posix_fadvise() 2008-01-27 19:51:40 +00:00
dsl 460b556c90 Move the prototype for do_posix_fadvise() somewhere useful. 2008-01-27 19:48:52 +00:00
pooka 67b0c1ab9c unlock vp in error branch 2008-01-27 19:19:42 +00:00
pooka e90887eb2e vrele2 is a goner 2008-01-27 19:08:07 +00:00
pooka fb3107360b Use vfs_subr.c from sys/kern. This brings differences in the vnode 
life cycle between rump and a real kernel to a minimum.
 2008-01-27 19:07:20 +00:00
rmind 4c26b5ddf3 - sys_shmget: size type must be size_t, not int. Should avoid possible 
problems with huge allocations.
- shmrealloc: use newshmni for calculation of new sz, shminfo.shmmni
  was a mistake.  Convert sz to size_t type.
 2008-01-27 18:37:50 +00:00
dsl 5da7cb61ba Remove NETBSD32_OFF_T_RETURN() netbsd32_lseek() now uses MI code to return 
all 64 bits of the new position.
 2008-01-27 17:37:40 +00:00
hannken ce34b417f9 uvm_swap_init(): Call VOP_OPEN() on swapdev_vp to make I/O through the 
swap device work with specnodes.

Ok: Andrew Doran <ad@netbsd.org>
 2008-01-27 17:18:09 +00:00
chris 84400d5df4 Fix arm cpu_in_cksum. It was walking the mbuf chain wrongly, and trying 
to read from the wrong memory address and so causing protection faults.

Tested with iperf, mounting nfs drives, and also the regress tool.
 2008-01-27 16:58:05 +00:00
chris e083edeb16 Force the last loop to always walk all the data in the mbuf chain. 
It appears that randomness was against me when I updated the arm version
as it crashes when used for real.  By forcing one full walk of the mbuf
chain the regression tool crashes, hinting that something is wrong.
 2008-01-27 16:49:13 +00:00
martin 29ef53c683 libc part of posix_fadvise changes. 
The new compatibility stub is ABI compatible to the old syscall, so no
need to bump libc minor.
 2008-01-27 16:18:22 +00:00
martin 2e87d89112 Implement new version of posix_fadvise as a stub callinig the real 
worker function, and compatibility stub doing the same with old argument
sturcture.
 2008-01-27 16:16:50 +00:00
martin a870c2e30e Regen (posix_fadvise got a pad arg) 2008-01-27 16:14:59 +00:00
martin 65badb72df Version the posix_fadvise syscall and add the missing "pad" argument, 
so that it does not only work on i386 (no padding) and 64bit archs (all
args padded to 64bit).
 2008-01-27 16:13:39 +00:00
joerg 7a9167c057 pkg_install-20080126 2008-01-27 14:11:27 +00:00
chris 67bda63986 Add an arm/conf/std.arm file which contains common arm options, initially 
just options CPU_IN_CKSUM.

Include std.arm in all arm platform std files.

This should reenable the asm in_cksum code for all arm platforms.

Also remove the now unused in_cksum_arm.S.
 2008-01-27 12:37:10 +00:00
kiyohara fa0b00252c Update and add some options/devices. 2008-01-27 06:23:38 +00:00
xtraeme 6efe384afc Remove references to check_uucp in daily.conf(5) and the defaults 
file... uucp was removed a while ago.
 2008-01-27 05:20:24 +00:00
jmcneill cdbea784c7 On resume, ensure that esa is powered up, and ensure we power the 
AC97 codec once the host controller is up and running.
 2008-01-27 01:57:03 +00:00
jmcneill d66b2b71eb Fixup PCI_COMMAND_STATUS_REG on attach. 2008-01-27 01:56:02 +00:00
dyoung ae1c071d6f Patch /dev/drvctl and drvctl(8) to let us suspend/resume device 
sub-trees, and to let us list the children of a device.

While I am here, add -p to the drvctl(8) usage message.
 2008-01-27 01:38:33 +00:00
christos 4d6010e3ba add openpam_attr.h 2008-01-27 01:30:18 +00:00
christos 9d58de27f0 Install a new include file. 2008-01-27 01:27:05 +00:00