David E. O'Brien [Mon, 25 Mar 2002 03:27:09 +0000 (03:27 +0000)]
Remove a FMT that was buried in here.
I have no idea why the original committer even used it (in its KERNFORMAT
spelling) as there is no a.out version.
David E. O'Brien [Sun, 24 Mar 2002 17:45:46 +0000 (17:45 +0000)]
Remove the setting of 'FMT'. We now do ELF by default, so only bother
to increase the lenght of the command line if needed. The setting of FMT
also gets in the cross bootstrapping way for new platforms.
Wilko Bulte [Sun, 24 Mar 2002 14:35:20 +0000 (14:35 +0000)]
plip has produced an error during attach on Alpha as long as I care
to remember. Nobody in the right sense of their mind would run IP on
the parallel port on Alpha so comment this one out.
Hide the ACPI counter probing printf behind bootverbose.
The conclusion is that this method really can tell the perfect from the
less than perfect ACPI counters.
It is in fact probably a bit more discriminative than that, but we
will rather condemn some otherwise perfect counters to the slightly
slower "-safe" version, than certify a counter as perfect which
will let us down later.
Many thanks to all the people who sent email reports!
Jeff Roberson [Sun, 24 Mar 2002 10:56:11 +0000 (10:56 +0000)]
Reset the cachefree statistics after draining the cache. This fixes a bug
where a sysctl within 20 seconds of a cache_drain could yield negative "USED"
counts.
Also, grab the uma_mtx while in the sysctl handler. This hadn't caused
problems yet because Giant is held all the time.
Bruce Evans [Sun, 24 Mar 2002 10:19:10 +0000 (10:19 +0000)]
Fixed some style bugs in the removal of __P(()). Continuation lines
were not outdented to preserve non-KNF lining up of code with parentheses.
Switch to KNF formatting.
Bruce Evans [Sun, 24 Mar 2002 09:34:04 +0000 (09:34 +0000)]
Fixed some style bugs in the removal of __P(()). The main ones were
not removing tabs before "__P((", and not outdenting continuation lines
to preserve non-KNF lining up of code with parentheses. Switch to KNF
formatting and/or rewrap the whole prototype in some cases.
__FBSDID, ANSIfy function definitions, staticize functions and globals,
don't initialize globals to 0, some reindentation of declarations, some
comments added or reworded.
Bruce Evans [Sun, 24 Mar 2002 05:09:11 +0000 (05:09 +0000)]
Fixed some style bugs in the removal of __P(()). The main ones were
not removing tabs before "__P((", and not outdenting continuation lines
to preserve non-KNF lining up of code with parentheses. Switch to KNF
formatting and/or rewrap the whole prototype in some cases.
Bruce Evans [Sun, 24 Mar 2002 04:35:23 +0000 (04:35 +0000)]
Fixed some style bugs in the removal of __P(()). Continuation lines
were not outdented to preserve non-KNF lining up of code with parentheses.
Switch to KNF formatting in some cases.
Bruce Evans [Sun, 24 Mar 2002 04:09:05 +0000 (04:09 +0000)]
Fixed some style bugs in the removal of __P(()). Continuation lines
were not outdented to preserve non-KNF lining up of code with parentheses.
Switch to KNF formatting.
Thomas Moestl [Sun, 24 Mar 2002 02:50:53 +0000 (02:50 +0000)]
Revamp the busdma implementation a bit:
- change the IOMMU support code so that it supports overcommittting the
available DVMA memory, while still allocating as lazily as possible.
This is achieved by limiting the preallocation, and deferring the
allocation to map load time when it fails. In the latter case, the
DVMA memory reserved for unloaded maps can be stolen to free up enough
memory for loading a map.
- allow NULL settings in the method tables, and search the parent tags
until an appropriate implementation is found. This allows to remove some
kluges in the old implementation.
Thomas Moestl [Sun, 24 Mar 2002 02:11:06 +0000 (02:11 +0000)]
Make the OpenFirmware interrupt mapping code more generic, to reduce
the bus-dependent code and to be able to support more systems. The core
of the new code is mostly obtained from NetBSD.
Kluge the interrupt routing methods of the psycho and apb drivers so
that an intline of 0 can be handled for now; real routing is still not
possible (all intline registers are preinitialized instead); this will
require a sparc64-specific adaption of the driver for generic PCI-PCI
bridges with a custom routing method to work right.
Thomas Moestl [Sat, 23 Mar 2002 20:42:23 +0000 (20:42 +0000)]
Add code to print the fault virtual address for uncorrectable DMA errors
caused by IOMMU misses to aid debugging. This will only work on
UltraSPARC-IIi and IIe.
Thomas Moestl [Sat, 23 Mar 2002 19:43:15 +0000 (19:43 +0000)]
In some cases, RX descriptors that are signalled to have been completed
by the hardware are still marked as owned. Handle this by installing a
timeout handler to collect this descriptor to avoid having received
packets remain unhandled until the next one arrives.
Thomas Moestl [Sat, 23 Mar 2002 19:37:11 +0000 (19:37 +0000)]
Assorted fixes:
- remove some useless code from the status change handler that was intended
to enable the the MII drivers for external phys; this is already done
during interface initialization, and the deleted code made some
assumptions about phy addresses that do not seem to hold true on newer
cards. This should get at least one of the two hmes of newer Netra t1
machines working.
- correct the interrupt resource allocation
- bump the number of RX descriptors, lower values cause promblems on some
machines
Automatically detect devices that do not support READ(6)/WRITE(6)
and upgrade to using 10 byte cdbs.
As far as I tested, this works efficiently for most of the
SBP-II/Firewire devices but most of the umass devices still need
ad-hoc work around because umass-sim doesn't return any SCSI errors.
A sysctl nob is also added for the last resort.
I hope we don't need DA_Q_NO_6_BYTE quirks anymore.
Install headers with -C. Ideally, these Makefiles should not need to
override the beforeinstall target at all, but this has proven difficult
to achieve.
Warner Losh [Sat, 23 Mar 2002 17:24:55 +0000 (17:24 +0000)]
Breath deep and take __P out of the system include files.
# This appears to not break X11, but I'm having problems compiling the
# glide part of the server with or without this patch, so I can't tell
# for sure.
Jake Burkholder [Sat, 23 Mar 2002 17:04:41 +0000 (17:04 +0000)]
Cleanup the trace back routine slightly. Print the leaf return value so
that traps inside of leaf functions are less confusing. Add a function
to print a non-symbolic trace of the user stack.