Try to honor BUS_DMA_COHERENT : if the flag is set, normally allocate memory
with malloc() or contigmalloc() as usual, but try to re-map the allocated
memory into a VA outside the KVA, non-cached, thus making the calls to
bus_dmamap_sync() for these buffers useless.
Hajimu UMEMOTO [Wed, 1 Mar 2006 16:13:17 +0000 (16:13 +0000)]
- Reduce needless DNS query by lookup only appropriate address
family. [1]
- Specify appropriate hints to getaddrinfo(3). [1]
- Obtain address family from peername in inet mode.
Søren Schmidt [Wed, 1 Mar 2006 07:48:41 +0000 (07:48 +0000)]
Work around the deadlock that occours when ATA waits for the taskqueue
to call back for completition and something else is holding the taskqueue
waiting for ATA to return data.
This should clear up the "semaphore timeout !! DANGER Will Robinson !!"
in most situations, and log "taskqueue timeout - completing request directly"
instead, with a delayed "WARNING - freeing taskqueue zombie request" when
the taskqueue finally calls us back with the now stale request.
(It would have been nice if there was a way to remove a scheduled item from
a taskqueue, but that is not currently implemented in the kernel).
A real fix for this is in the works but wont make it to 6.1RELEASE
Scott Long [Wed, 1 Mar 2006 07:24:39 +0000 (07:24 +0000)]
Big update to the iir driver:
- Don't use a common buffer in the softc to store per-command data. Reserve
a buffer in the command itself.
- Don't allocate DMA memory for the kernel command structures when all you
really need is DMA memory for the scratch buffer embedded in them. Instead
allocate a slab for the scratch buffers and divide it up as needed.
- Call bus_dmamap_unload() at the completion of commands.
- Preserve and clear the CAM CCB status flags at completion.
- Reorder some low-level command operations to try to close races.
- Limit the simq to 32 commands for now. There are some serious problems
with the driver under load that are not well understood, so keeping the
simq lower helps avoid this. It has been tested at a higher value, but
this is a safe value that doesn't show much performance degredation.
These changes allow the driver to work reliably with >4GB of memory on i386
and amd64, and also work around deadlocks seen under very high load in
certain situations. The work-around is far from ideal, but without and
documentation it is hard to know what the right fix is.
David Xu [Wed, 1 Mar 2006 06:37:34 +0000 (06:37 +0000)]
Bring in my initial version of POSIX realtime extension library.
Current the library implements mqueue, timer and aio with SIGEV_THREAD
notification supported.
David Xu [Wed, 1 Mar 2006 06:29:34 +0000 (06:29 +0000)]
Let kernel POSIX timer code and mqueue code to use integer as a resource
handle, the timer_t and mqd_t types will be a pointer which userland
will define it.
Maksim Yevmenkin [Tue, 28 Feb 2006 23:46:23 +0000 (23:46 +0000)]
Integrate kbdmux(4) into syscons(4) and kbd code.
By default syscons(4) will look for the kbdmux(4) keyboard first, and then,
if not found, look for any keyboard.
Current kbd code is modified so if kbdmux(4) is the current keyboard, all
new keyboards are automatically added to the kbdmux(4).
Switch to kbdmux(4) can be done at boot time, by loading kbdmux module at
the loader prompt, or at runtime, by kldload'ing the kbdmux module and
releasing current active keyboard.
If, for whatever reason, kbdmux(4) is not required/desired then just do
not load it and everything should work as before. It is also possible to
kldunload kbdmux at runtime and syscons(4) will automatically switch to
the first available keyboard.
No response from: freebsd-current@
MFC after: 1 day
Andre Oppermann [Tue, 28 Feb 2006 23:05:59 +0000 (23:05 +0000)]
Rework TCP window scaling (RFC1323) to properly scale the send window
right from the beginning and partly clean up the differences in handling
between SYN_SENT and SYN_RCVD (syncache).
Further changes to this code to come. This is a first incremental step
to a general overhaul and streamlining of the TCP code.
John Baldwin [Tue, 28 Feb 2006 22:24:55 +0000 (22:24 +0000)]
Rework how we wire up interrupt sources to CPUs:
- Throw out all of the logical APIC ID stuff. The Intel docs are somewhat
ambiguous, but it seems that the "flat" cluster model we are currently
using is only supported on Pentium and P6 family CPUs. The other
"hierarchy" cluster model that is supported on all Intel CPUs with
local APICs is severely underdocumented. For example, it's not clear
if the OS needs to glean the topology of the APIC hierarchy from
somewhere (neither ACPI nor MP Table include it) and setup the logical
clusters based on the physical hierarchy or not. Not only that, but on
certain Intel chipsets, even though there were 4 CPUs in a logical
cluster, all the interrupts were only sent to one CPU anyway.
- We now bind interrupts to individual CPUs using physical addressing via
the local APIC IDs. This code has also moved out of the ioapic PIC
driver and into the common interrupt source code so that it can be
shared with MSI interrupt sources since MSI is addressed to APICs the
same way that I/O APIC pins are.
- Interrupt source classes grow a new method pic_assign_cpu() to bind an
interrupt source to a specific local APIC ID.
- The SMP code now tells the interrupt code which CPUs are avaiable to
handle interrupts in a simpler and more intuitive manner. For one thing,
it means we could now choose to not route interrupts to HT cores if we
wanted to (this code is currently in place in fact, but under an #if 0
for now).
- For now we simply do static round-robin of IRQs to CPUs when the first
interrupt handler just as before, with the change that IRQs are now
bound to individual CPUs rather than groups of up to 4 CPUs.
- Because the IRQ to CPU mapping has now been moved up a layer, it would
be easier to manage this mapping from higher levels. For example, we
could allow drivers to specify a CPU affinity map for their interrupts,
or we could allow a userland tool to bind IRQs to specific CPUs.
The MFC is tentative, but I want to see if this fixes problems some folks
had with UP APIC kernels on 6.0 on SMP machines (an SMP kernel would work
fine, but a UP APIC kernel (such as GENERIC in RELENG_6) would lose
interrupts).
John Baldwin [Tue, 28 Feb 2006 21:49:33 +0000 (21:49 +0000)]
Fix numerous warnings. Aside from menu items in system.c and menu.c
this now compiles on i386 with WARNS?= 3. Most of the fixes included
adding missing 'static' keywords to internal functions, using fully-defined
terminators in statically defined arrays of structs, and various
signed vs unsigned mismatches. Also G/C'd unused configSecurity()
function.
John Baldwin [Tue, 28 Feb 2006 21:12:16 +0000 (21:12 +0000)]
Use some helper macros for the device table so it can fit in 80 columns and
to allow us to use default values in some entries to quiet warnings. This
table is actually readable again now.
John Baldwin [Tue, 28 Feb 2006 20:29:43 +0000 (20:29 +0000)]
- Autogenerate a menu containing a list of countries and keymaps supported
by syscons.
- If we are running as init, popup the country menu before the main menu.
If a non-default country is chosen, then a second menu is brought up
to let the user choose a keymap. By default the default keymap for
the country that was selected is highlighted. If the user chooses the
default country, then the default keymap is just assumed and the user
is not presented with the keymap menu. Currently the default country
is set to "United States" except for PC98 which assumes "Japan".
PR: bin/93853
Submitted by: Seth Kingsley sethk at magnesium dot net
MFC after: 3 days
John Baldwin [Tue, 28 Feb 2006 20:11:30 +0000 (20:11 +0000)]
Allow PHOLD()'s of curproc even if P_WEXIT is set. Normally we don't want
to allow PHOLD()'s of processes that have P_WEXIT set as once that flag
is set we aren't guaranteed to block in exit1() waiting for the PRELE()
(we might already be past the wait). However, curproc is a bit of a
special case. By the time P_WEXIT is set, the process is single-threaded,
so the only thread for which can do a PHOLD(curproc) is the thread
executing in exit1(). The fact that this thread is executing ensures
that the process won't go away before the current hold is released via
PRELE(). This fixes some panics due to kicking off softupdate operations
inside of exit1() after the recent PHOLD changes to fix ptrace/procfs vs
exit races.
Paul Saab [Tue, 28 Feb 2006 19:39:18 +0000 (19:39 +0000)]
Fix 32bit sendfile by implementing kern_sendfile so that it takes
the header and trailers as iovec arguments instead of copying them
in inside of sendfile.
Luigi Rizzo [Tue, 28 Feb 2006 08:02:28 +0000 (08:02 +0000)]
use standard mode instead of 500 for /sbin/init.
As discussed on -current, there is no sensitive info in /sbin/init
to prevent reading it from non-privileged users, nor any reason to
remove the 'x' bit as the first thing the program does is check the
uid and exit if it is not run by root.
Instead (and this is why i make the change), mode 500 prevents
operation when exporting the partition without -maproot=0 to diskless
clients.
All previuos releases are affected by the same problem, so a merge
to RELENG_6 at least would be appropriate (after proper re@ approval
of course).
Matt Jacob [Tue, 28 Feb 2006 07:44:50 +0000 (07:44 +0000)]
Fix mpt_reset to try mpt_hard_reset more than once, and to try
mpt_soft_reset more than once. And to wait for MPT_DB_STATE_READY
twice. I mean, this is crucial- give the IOC a chance to get
ready.
If mpt_reset is called to reinit things, and we succeed, make
sure to re-enable interrupts. This is what has mostly led to
system lockup after having to hard reset the chip. Also, if
we think that interrupts aren't function in mpt_cam_timeout,
for goodness sake, turn them on again.
In read_cfg_header, return distinguishing errnos so the caller
can decide what's an error. It's *not* an error to fail to
read a RAID page from a non-RAID capable device like the FC929X.
Some whitespace fixes (removing spaces from ends of lines).
Maksim Yevmenkin [Tue, 28 Feb 2006 03:34:06 +0000 (03:34 +0000)]
Make ukbd(4) to not delay break scancodes in "raw" mode.
Developed with: Norbert Koch < NKoch at demig dot de >
No response from: freebsd-current@
Tested by: Norbert Koch < NKoch at demig dot de >
MFC after: 1 day
Andrew Thompson [Tue, 28 Feb 2006 00:13:24 +0000 (00:13 +0000)]
Fix up the Bridge Identifier field in the BPDU packet.
- use the cu_bridge_id rather than the cu_rootid for the bridge address [1]
- the memcmp return value is not signed so the wrong interface may have been
selected
- fix up the calculation of sc_bridge_id
Extend kldunload(8) functionality and fix minor problems:
o multiple modules can be unloaded at once (specified either by id or be
module name)
o exit with EX_USAGE after usage() is called.
o remove unused variables, since we keep command line flags as bitmask,
in 'opt'.
o 'kldload -n ...' does nothing. Add comment to this options.
Additionally:
o Update manual page to conform new functionality.
o Increace WARNS to 6. Because we can.
Daniel Eischen [Mon, 27 Feb 2006 22:08:16 +0000 (22:08 +0000)]
Correct a comment.
Staticize two tables thare are not visible in <resolv.h>
and which are also local in Solaris' libresolv.
Remove two functions that are not referenced in libc nor
anywhere else I can find, not visible in <resolv.h> and
which are also local in Solaris libresolv.
David Malone [Mon, 27 Feb 2006 19:13:47 +0000 (19:13 +0000)]
Avoid moving onto a new line while testing if there is a syslog priority
in the buffer. This isn't exactly the patch that Stephen submitted, but
is based on one of his suggestions.
This patch fixes a problem, which exists if you have IPSEC in your kernel
and want to have crypto support loaded as KLD. By moving zlib to separate
module and adding MODULE_DEPEND directives, it is possible to use such
configuration without complication. Otherwise, since IPSEC is linked with
zlib (just like crypto.ko) you'll get following error:
interface zlib.1 already present in the KLD 'kernel'!
Yaroslav Tykhiy [Mon, 27 Feb 2006 12:04:13 +0000 (12:04 +0000)]
CODA_COMPAT_5 may not be defined unconditionally in the coda5 module.
Otherwise a kernel build would break in the coda5 module if the main
kernel conf file enabled CODA_COMPAT_5, too. Redefined symbols are
strictly disallowed by -Werror.
To overcome this issue, introduce a different symbol indicating coda5
build, CODA5_MODULE, and translate it to CODA_COMPAT_5 appropriately
in /sys/coda/coda.h.
Matt Jacob [Sun, 26 Feb 2006 22:50:14 +0000 (22:50 +0000)]
Remove the ill-considered effect of using the type definitions as
distributed by LSI-Logic. For FreeBSD, just use the posix defines
instead of trying to figure out how wide an int is. Apologies to all.
Warner Losh [Sun, 26 Feb 2006 17:45:35 +0000 (17:45 +0000)]
o Bump date (thanks ru)
o Also, fdisk_pc98 appears to support -i, so add it to the man page.
o Ditto -v
o Change the name from PC partition table maintenance program to NEC PC-98x1
partition table maintenance program, since it is possible to have a DOS MBR
formatted disk on FreeBSD/pc98 now (although we don't install any tools for
this).
Xin LI [Sun, 26 Feb 2006 14:03:13 +0000 (14:03 +0000)]
Provide SWIDTH information so that applications like tcsh(1) can properly
handle with zh_CN.GBK locale, which is quite common for users who uses
Simplified Chinese.
David Xu [Sun, 26 Feb 2006 12:56:23 +0000 (12:56 +0000)]
1. Remove aio entry from lists earlier in aio_free_entry,
so other threads can not see it if we unlock the proc
lock (this can happen in knlist_delete). Don't do wakeup,
it is not necessary.
2. Decrease kaio_buffer_count in biohelper rather than
doing it in aio_bio_done_notify.
3. In aio_bio_done_notify, don't send notification if KAIO_RUNDOWN
was set, because the process is already in single thread mode.
4. Use assignment to initialize aiothreadflags.
5. AIOCBLIST_RUNDOWN is not useful, axe the code using it.
Matt Jacob [Sun, 26 Feb 2006 07:46:09 +0000 (07:46 +0000)]
Shorten the time for waiting for TMF commands to complete- let's
not hang the system for 5 seconds. If a TMF doesn't complete within,
oh, say 500ms, that's enough.
Put in a printout to catch mpt_recover_commands being activated with
no commands.
Matt Jacob [Sun, 26 Feb 2006 07:44:31 +0000 (07:44 +0000)]
a) Delay for port enable to succeed should be 30 seconds (at least) for
*both* SAS and FC, not just SAS.
b) Don't tell the chip we want it to do FIFO signalling if we actually
don't set up the address where the FIFO signal is supposed to be written
(oops).
Warner Losh [Sun, 26 Feb 2006 07:37:10 +0000 (07:37 +0000)]
Fixes and new features:
o Merge the -I switch from i386 verison to initialize the partition table
to use the maximum amount possible for a single FreeBSD table.
o Improve warning when the geom method fails (which I think it
always will until geom_pc98 is updated to respond to this ctl message)
o when writing out the boot sector, we have to write out a minimum of
1024 bytes or the sector size. This is different than the i386 case where
we need to write out a minimum of 512 bytes (which is also the minimum
sector size). We already handle this difference on reading, but didn't
in writing, so attempting to write a new partition table would fail.
o Add MID to the -s output, since pc98 users are likely interested in
both of these parameters.
# I can now initialize disks on my pc98 machine either by -I or by
# manually entering the parameters. I don't know if fdisk -B works or not,
# since I'm not willing to risk my only working boot disk to test it..
projects / FreeBSD / FreeBSD.git / log