Unfortunately there was a typo in one of the patches I sent,
(The one now committed to the agpgart tree).
It may cause a bus error on i810 type hardware.
Signed-off-by: Thomas Hellstrom <thomas@tungstengraphics.com>
Signed-off-by: Dave Jones <davej@redhat.com>
On the G965, the GTT size may be larger than is required to cover the
aperture. (In fact, on all hardware we've seen, the GTT is 512KB to the
aperture's 256MB). A previous commit forced the aperture size to 512MB on
G965 to match GTT, which would likely result in hangs at best if users
tried to rely on agpgart's aperture size information. Instead, we use the
resource length for the aperture size and the system's reported GTT size
when available for the GTT size.
Because the MSAC registers which had been read for aperture size detection
on i9xx chips just cause a change in the resource size, we can use generic
code for aperture detection on all i9xx.
Signed-off-by: Eric Anholt <eric@anholt.net>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Dave Jones <davej@redhat.com>
This patch is to speed up flipping of pages in and out of the AGP aperture as
needed by the new drm memory manager.
A number of global cache flushes are removed as well as some PCI posting flushes.
The following guidelines have been used:
1) Memory that is only mapped uncached and that has been subject to a global
cache flush after the mapping was changed to uncached does not need any more
cache flushes. Neither before binding to the aperture nor after unbinding.
2) Only do one PCI posting flush after a sequence of writes modifying page
entries in the GATT.
Signed-off-by: Thomas Hellstrom <thomas@tungstengraphics.com>
Signed-off-by: Dave Jones <davej@redhat.com>
Not all graphic page remappers support physical addresses over the 4GB
mark for remapping, so while some do (the AMD64 GART always did, and I
just fixed the i965 to do so properly), we're safest off just forcing
GFP_DMA32 allocations to make sure graphics pages get allocated in the
low 32-bit address space by default.
AGP sub-drivers that really care, and can do better, could just choose
to implement their own allocator (or we could add another "64-bit safe"
default allocator for their use), but quite frankly, you're not likely
to care in practice.
So for now, this trivial change means that we won't be allocating pages
that we can't map correctly by mistake on x86-64.
[ On traditional 32-bit x86, this could never happen, because GFP_KERNEL
would never allocate any highmem memory anyway ]
Acked-by: Andi Kleen <ak@suse.de>
Acked-by: Dave Jones <davej@redhat.com>
Cc: Eric Anholt <eric@anholt.net>
Cc: Keith Packard <keithp@keithp.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
This introduces a i965-specific "mask_memory()" function that knows
about the extended physical addresses that the i965 supports. This
allows us to correctly map in physical memory in the >4GB range into the
GTT.
Also simplify/clean-up the i965 case for the aperture sizing by just
returning the fixed 512kB size from "fetch_size()". We don't really
care that not all of the aperture may be visible - the only thing that
cares about the aperture size is the Intel "stolen memory" calculation,
which depends on the fixed size.
Cc: Keith Packard <keithp@keithp.com>
Cc: Eric Anholt <eric@anholt.net>
Cc: Dave Jones <davej@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
Here's a very small diff for 945GM support for agpgart.
Patch against 2.6.15.
From: Alan Hourihane <alanh@fairlite.demon.co.uk>
Signed-off-by: Dave Jones <davej@redhat.com>
AGP allocation/deallocation is suffering major performance issues due to
the nature of global_flush_tlb() being called on every change_page_attr()
call.
For small allocations this isn't really seen, but when you start allocating
50000 pages of AGP space, for say, texture memory, then things can take
seconds to complete.
In some cases the situation is doubled or even quadrupled in the time due
to SMP, or a deallocation, then a new reallocation. I've had a case of
upto 20 seconds wait time to deallocate and reallocate AGP space.
This patch fixes the problem by making it the caller's responsibility to
call global_flush_tlb(), and so removes it from every instance of mapping a
page into AGP space until the time that all change_page_attr() changes are
done.
Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
From: Laurent Riffard <laurent.riffard@free.fr>
This updates .owner field of struct pci_driver.
This allows SYSFS to create the symlink from the driver to the module which
provides it.
$ tree /sys/bus/pci/drivers/agpgart-via/
/sys/bus/pci/drivers/agpgart-via/
|-- 0000:00:00.0 -> ../../../../devices/pci0000:00/0000:00:00.0
|-- bind
|-- module -> ../../../../module/via_agp
|-- new_id
`-- unbind
Signed-off-by: Laurent Riffard <laurent.riffard@free.fr>
Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
When leaving S3 state, the AGP bridge may not have all PCI configuration
registers set in the same way as they were at boot. This should be fixed
by pci_restore_state - however, the APBASE register cannot be set to
conflict with the APSIZE register. If APSIZE is larger than it was before
suspend, pci_restore_state will not restore APBASE correctly. The attached
patch adds an extra item to the agp_bridge_data structure and uses it to
store the value of APBASE. On resume, this is then written after APSIZE
has been set. This patch only touches the path used for Intel chipsets
without integrated graphics, and may need to be extended to work with the
others.
Without this patch, I get the symptoms described in bug 4921 - APBASE ends
up overlapping various PCI devices, and as a result they fail to work after
resume.
Signed-off-by: Matthew Garrett <mjg59@srcf.ucam.org>
Acked-by: Dave Jones <davej@redhat.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
When Linux is running on the Xen virtual machine monitor, physical
addresses are virtualised and cannot be directly referenced by the AGP
GART. This patch fixes the GART driver for Xen by adding a layer of
abstraction between physical addresses and 'GART addresses'.
Architecture-specific functions are also defined for allocating and freeing
the GATT. Xen requires this to ensure that table really is contiguous from
the point of view of the GART.
These extra interface functions are defined as 'no-ops' for all existing
architectures that use the GART driver.
Signed-off-by: Keir Fraser <keir@xensource.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Dave Jones <davej@redhat.com>
Attached is a small patch for i945G support against 2.6.11.11.
From: Alan Hourihane <alanh@fairlite.demon.co.uk>
Signed-off-by: Dave Jones <davej@redhat.com>
Initial git repository build. I'm not bothering with the full history,
even though we have it. We can create a separate "historical" git
archive of that later if we want to, and in the meantime it's about
3.2GB when imported into git - space that would just make the early
git days unnecessarily complicated, when we don't have a lot of good
infrastructure for it.
Let it rip!
Thomas Hellstrom