More code needs to be guarded by if (uss32) { ... }. Without this
patch, if (buf32..) may be checking uninitialized data and may result
in spurious ENOMEMs.
Signed-off-by: Gordon Jin <gordon.jin@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Rename SN local function move_irq to sn_move_irq to remove conflict
with global move_irq.  Needed since ashok.raj moved move_irq() from
iosapic.c to irq.c.
Signed-off-by: Keith Owens <kaos@sgi.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Trivial patch to quiet messages about corrected errors.
Signed-off-by: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

I'm hitting this problem because the module I want to load was
compiled with -g, so the filesize is absolutely huge.  kernel/module.c
does a vmalloc() to inhale the entire file, and then two calls to
module_alloc() (which calls vmalloc() on ia64) to load the 'init' and
'core' sections. The 'init' is small and slips into a gap early in the
vmalloc playspace, while the 'core' area is allocated after the huge
area that was allocated for the inhaled copy of the whole file.

I made a one coding change to Jean-Marc's version, adding a check to
see whether the init/core sections are close enough together for the
PCREL21B to reach (they almost always are).

I've kept the test that there are no jumps from core to init (though
the message that is printed is almost as unhelpful as the one that you
get when you statically link a module into the kernel that has calls
to the discarded .exit section :-)

This patch stalled out before on the question of whether a PLT was
overkill for a section-to-section branch, and whether it would be
better to use relaxation techniques to extend the range.  Doing that
might be more elegant in some way, but it needs a whole heap of extra
code (to count how many relaxed branches need to be added and
allocated them as needed).  This seems a lot of extra code that is
only ever going to be exercised by maniacs like me with 49MB .ko
files.

I fixed up the comments to be a little more heplful.

Signed-off-by: <tony.luck@intel.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

The Altix console driver has been replaced in Andrew's tree by a new
one that uses the serial core API.  Update the early_console_setup
routine to use the new config option so that it's ready when the
driver hits Linus' tree.  This will also fix a build bug since the old
driver hasn't been updated.
Signed-off-by: Jesse Barnes <jbarnes@sgi.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

The patch to move the current pointer for init_task from region 7 to
region 5 inadvertendly caused ar.k4 (CURRENT_STACK) to be initialized
improperly for the bootstrap processor.  Fix this by initializing it
to -1.
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

The code allocated all 96 stacked registers, for no apparent reason.

Also, fix ia64_pal_call_static() to switch RSE to enforced-lazy, LE
mode before the PAL-call.
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Change ia64_switch_mode_phys() and ia64_switch_mode_virt() to preserve
the virtual sp/bsp and update the call-sites accordingly.  This avoids
problems with the init_task pointer which lives in region 5 now.
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Gcc 3.4 has a new DFA scheduler for ia64. Unfortunately, it is not
well tested for Itanium 1. When I tried to use gcc 3.4 to compile
2.6 kernel configured for Merced, gcc failed in many places:

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15598
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15653
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=15655
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16130
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16142
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16143

Some of them have been fixed. But I am afraid many more remain. I don't
know how many Itanium 1 machines left in production use. If people want
to optimize kernel for Itanium 1, they can stick with the older gccs.
Or they should upgrade their machines. I think the gcc developer's time
should be better spent on somewhere else. I have been using this patch
for a while, kernel seems to work OK.
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Define cpu_logical_id() even when !SMP.  I added uses of this in some
iosapic printk's, which broke the UP build.
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

call_pernode_memory() is rounding memory regions to page boundaries,
but in the non-NUMA case, it's incorrectly passing the rounded start and
the original length.  Any reserved memory regions sitting beyond a
memory region that isn't nicely aligned is potentially at risk for
getting freed and trashed.  This is a fairly nasty bug that can bite in
very hard to find ways.  In my case, the EFI memory map was getting
overwritten, causing the ACPI methods to access chipset registers
cachable instead of uncachable.  It could potentially be causing some of
the random hangs and strange behavior we're seeing lately.
Signed-off-by: Alex Williamson <alex.williamson@hp.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

This update fixes the FAT default NLS options and trims some others.
Signed-off-by: Jesse Barnes <jbarnes@sgi.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Introduce iosapic_{read,write,eoi} to hide the IOSAPIC REG_SELECT
stuff.  i386 and x86_64 already do this for io_apic_{read,write}.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Altix needs to use 16MB granules to prevent attribute aliasing in low
memory space.  This patch makes the kernel configuration system
enforce that requirement.
Signed-off-by: Jesse Barnes <jbarnes@sgi.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

The attached patch contains some cleanups to the bte code and introduces
the ability to excercise bte interfaces on other nodes.  This is an SGI
Altix specific piece of hardware.

Signed-off by: Robin Holt
Signed-off by: Jesse Barnes
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Allow consoles to register early if they're capable of early output.
If one of them registers successfully, set CPU 0 online so that
printk() will actually call the console drivers.
Signed-off-by: Jesse Barnes <jbarnes@sgi.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

In base kernel 2.6.7, there is bug in function hugetlb_prefault().  The
call to free_huge_page() is incorrect and should be replaced with a call
to page_cache_release().
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

This patch moves the pfm_ovfl_arg_t structure from the kernel stack
into the pfm_context_t structure. This minimizes kernel stack space
consumption.
Signed-off-by: Stephane Eranian <eranian@hpl.hp.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

This patch moves move_irq() from ioaspic.c to irq.c to make this common
for ia64 subarches.
Signed-off-by: Ashok Raj <ashok.raj@intel.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

- Handle the overlapping mmaps in the MAP_FIXED case
- Handle unmaps correctly i.e. call __ia32_delete_pp_range()
- There are now three clear entry points into the pp manipulation code
        - ia32_set_pp
        - ia32_unset_pp
        - ia32_compare_pp
- Rename internal functions to __ia32_*pp_*
- Grab mmap_sem just once at the main entry point
Signed-off-by: Gordon Jin <gordon.jin@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

A couple of months ago Hidetoshi Seto from Fujitsu proposed
a patch to provide similar switching between interrupt mode
and polling mode for corrected platform events (CPE) as we
have for processor events (CMC) [with the obvious difference
that not all platforms support an interrupt for CPE].

I dusted it off and made a couple of very minor cleanups (which
Seto-san checked out last night on his test setup and confirmed
still passes).
Signed-off-by: Tony Luck <tony.luck@intel.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Signed-off-by: Patrick Gefre <pfg@sgi.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

Without this fix, you'll get unresolved references to
sys_rt_sigaction().

into tiger.hpl.hp.com:/data1/bk/lia64/to-linus-2.5

Small patch to update the sn2_defconfig file.  I just did a 'make oldconfig' 
with the current sn2_defconfig and removed 8250 support.
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

The SN prom supports fine grained error handling features, the OS needs
to tell the prom if the OS expects to use these platform specific
features.
Signed-off-by: Keith Owens <kaos@sgi.com>
Signed-off-by: David Mosberger <davidm@hpl.hp.com>

into kernel.bkbits.net:/home/davem/sparc-2.6

I noticed that insert_resource() incorrectly handles the case of an
existing parent resource with the same ending address as a newly added
child.  This results in incorrect nesting, like the following:

	# cat /proc/ioports
	<snip>
	002f0000-002fffff : PCI Bus #48
	  00200000-002fffff : /pci@800000020000003
	</snip>
Signed-off-by: John Rose <johnrose@austin.ibm.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

Make sure that number of sectors != 0 for in/out command before sending
it to drive.  Remove no longer needed checks from flagged_* PIO handlers.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@elka.pw.edu.pl>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

PIO handlers under CONFIG_IDE_TASKFILE_IO=n are never used for bio
based requests (rq->bio is always NULL) so we can use rq->buffer
directly instead of calling ide_[un]map_buffer().
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@elka.pw.edu.pl>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

- merge status checking code for rq->current_nr_sectors
  and !rq->current_nr_sectors cases
- remove !rq->bio check as it is always true
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@elka.pw.edu.pl>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

These handlers are nowadays used only for REQ_DRIVE_TASKFILE
requests (rq->bio is always NULL) which aren't retried et all so
remove code 'rewinding' rq->current_nr_sectors and some FIXMEs.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@elka.pw.edu.pl>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

If REQ_DRIVE_* request fails ide_end_drive_cmd() should be called
for it not ->end_request().  This was broken by 2.6.5, fix it.
Signed-off-by: Bartlomiej Zolnierkiewicz <bzolnier@elka.pw.edu.pl>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>