The current radeon driver initialization routines, when using KMS, are written
so that the IRQ installation routine is called before initializing the WB buffer
and the CP rings. With some ASICs, though, the IRQ routine tries to access the
GFX_INDEX ring causing a call to RREG32 with the value of -1 in
radeon_fence_read. This, in turn causes the system to completely hang with some
cards, requiring a hard reset.
A call stack that can cause such a hang looks like this (using rv515 ASIC for the
example here):
* rv515_init (rv515.c)
* radeon_irq_kms_init (radeon_irq_kms.c)
* drm_irq_install (drm_irq.c)
* radeon_driver_irq_preinstall_kms (radeon_irq_kms.c)
* rs600_irq_process (rs600.c)
* radeon_fence_process - due to SW interrupt (radeon_fence.c)
* radeon_fence_read (radeon_fence.c)
* hang due to RREG32(-1)
The patch moves the IRQ installation to the card startup routine, after the ring
has been initialized, but before the IRQ has been set. This fixes the issue, but
requires a check to see if the IRQ is already installed, as is the case in the
system resume codepath.
I have tested the patch on three machines using the rv515, the rv770 and the
evergreen ASIC. They worked without issues.
This seems to be a known issue and has been reported on several bug tracking
sites by various distributions (see links below). Most of reports recommend
booting the system with KMS disabled and then enabling KMS by reloading the
radeon module. For some reason, this was indeed a usable workaround, however,
UMS is now deprecated and disabled by default.
Bug reports:
https://bugzilla.redhat.com/show_bug.cgi?id=845745https://bugs.launchpad.net/ubuntu/+source/linux/+bug/561789https://bbs.archlinux.org/viewtopic.php?id=156964
Signed-off-by: Adis Hamzić <adis@hamzadis.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
After refactoring the _cs logic, we ended up with many
macros and constants that #define the same thing.
Clean'em up.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch eliminates ASIC-specific ***_cs_packet_next_reloc
functions and hooks up the new common function.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This function is not limited to r100, but it can dump a
(raw) packet for any ASIC. Rename it accordingly and move
its declaration to radeon.h
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We now have a common radeon_cs_packet_parse function
that is good for all ASICs. Hook it up and eliminate
ASIC-specific versions.
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Pull drm merge (part 1) from Dave Airlie:
"So first of all my tree and uapi stuff has a conflict mess, its my
fault as the nouveau stuff didn't hit -next as were trying to rebase
regressions out of it before we merged.
Highlights:
- SH mobile modesetting driver and associated helpers
- some DRM core documentation
- i915 modesetting rework, haswell hdmi, haswell and vlv fixes, write
combined pte writing, ilk rc6 support,
- nouveau: major driver rework into a hw core driver, makes features
like SLI a lot saner to implement,
- psb: add eDP/DP support for Cedarview
- radeon: 2 layer page tables, async VM pte updates, better PLL
selection for > 2 screens, better ACPI interactions
The rest is general grab bag of fixes.
So why part 1? well I have the exynos pull req which came in a bit
late but was waiting for me to do something they shouldn't have and it
looks fairly safe, and David Howells has some more header cleanups
he'd like me to pull, that seem like a good idea, but I'd like to get
this merge out of the way so -next dosen't get blocked."
Tons of conflicts mostly due to silly include line changes, but mostly
mindless. A few other small semantic conflicts too, noted from Dave's
pre-merged branch.
* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (447 commits)
drm/nv98/crypt: fix fuc build with latest envyas
drm/nouveau/devinit: fixup various issues with subdev ctor/init ordering
drm/nv41/vm: fix and enable use of "real" pciegart
drm/nv44/vm: fix and enable use of "real" pciegart
drm/nv04/dmaobj: fixup vm target handling in preparation for nv4x pcie
drm/nouveau: store supported dma mask in vmmgr
drm/nvc0/ibus: initial implementation of subdev
drm/nouveau/therm: add support for fan-control modes
drm/nouveau/hwmon: rename pwm0* to pmw1* to follow hwmon's rules
drm/nouveau/therm: calculate the pwm divisor on nv50+
drm/nouveau/fan: rewrite the fan tachometer driver to get more precision, faster
drm/nouveau/therm: move thermal-related functions to the therm subdev
drm/nouveau/bios: parse the pwm divisor from the perf table
drm/nouveau/therm: use the EXTDEV table to detect i2c monitoring devices
drm/nouveau/therm: rework thermal table parsing
drm/nouveau/gpio: expose the PWM/TOGGLE parameter found in the gpio vbios table
drm/nouveau: fix pm initialization order
drm/nouveau/bios: check that fixed tvdac gpio data is valid before using it
drm/nouveau: log channel debug/error messages from client object rather than drm client
drm/nouveau: have drm debugging macros build on top of core macros
...
Convert #include "..." to #include <path/...> in drivers/gpu/.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Dave Airlie <airlied@redhat.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Dave Jones <davej@redhat.com>
Let's allow GCC to optimize better.
This exposed some five unused functions, but this patch doesn't remove them.
Signed-off-by: Lauri Kasanen <cand@gmx.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Making it easier to control when it is executed.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
The IB pool is in gart memory, so it is completely
superfluous to unpin / repin it on suspend / resume.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
No need to malloc it any more.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Christian König <deathsimple@vodafone.de>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Since it is now identical to r100_gpu_is_lockup.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Nothing chipset or ring specific with it,
so also move it to radon_ring.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
It isn't chipset specific, so it makes no sense
to have that inside r100.c.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Removing all the different error messages and
having just one standard behaviour over all
chipset generations.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
It makes no sense at all to have more than one flag.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
tidy up the radeon_asic struct.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König<christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Each ring type may need a different variant.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Christian König<christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
If accel is not working many subsystem such as the ib pool might not be
initialized properly that can lead to segfault inside kernel when cs
ioctl is call with non working acceleration. To avoid this make sure
the accel working flag is false when an error in GPU startup happen and
return EBUSY from cs ioctl if accel is not working.
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Virtual address space are per drm client (opener of /dev/drm).
Client are in charge of virtual address space, they need to
map bo into it by calling DRM_RADEON_GEM_VA ioctl.
First 16M of virtual address space is reserved by the kernel.
Once using 2 level page table we should be able to have a small
vram memory footprint for each pt (there would be one pt for all
gart, one for all vram and then one first level for each virtual
address space).
Plan include using the sub allocator for a common vm page table
area and using memcpy to copy vm page table in & out. Or use
a gart object and copy things in & out using dma.
v2: agd5f fixes:
- Add vram base offset for vram pages. The GPU physical address of a
vram page is FB_OFFSET + page offset. FB_OFFSET is 0 on discrete
cards and the physical bus address of the stolen memory on
integrated chips.
- VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR covers all vmid's >= 1
v3: agd5f:
- integrate with the semaphore/multi-ring stuff
v4:
- rebase on top ttm dma & multi-ring stuff
- userspace is now in charge of the address space
- no more specific cs vm ioctl, instead cs ioctl has a new
chunk
v5:
- properly handle mem == NULL case from move_notify callback
- fix the vm cleanup path
v6:
- fix update of page table to only happen on valid mem placement
v7:
- add tlb flush for each vm context
- add flags to define mapping property (readable, writeable, snooped)
- make ring id implicit from ib->fence->ring, up to each asic callback
to then do ring specific scheduling if vm ib scheduling function
v8:
- add query for ib limit and kernel reserved virtual space
- rename vm->size to max_pfn (maximum number of page)
- update gem_va ioctl to also allow unmap operation
- bump kernel version to allow userspace to query for vm support
v9:
- rebuild page table only when bind and incrementaly depending
on bo referenced by cs and that have been moved
- allow virtual address space to grow
- use sa allocator for vram page table
- return invalid when querying vm limit on non cayman GPU
- dump vm fault register on lockup
v10: agd5f:
- Move the vm schedule_ib callback to a standalone function, remove
the callback and use the existing ib_execute callback for VM IBs.
v11:
- rebase on top of lastest Linus
v12: agd5f:
- remove spurious backslash
- set IB vm_id to 0 in radeon_ib_get()
v13: agd5f:
- fix handling of RADEON_CHUNK_ID_FLAGS
v14:
- fix va destruction
- fix suspend resume
- forbid bo to have several different va in same vm
v15:
- rebase
v16:
- cleanup left over of vm init/fini
v17: agd5f:
- cs checker
v18: agd5f:
- reworks the CS ioctl to better support multiple rings and
VM. Rather than adding a new chunk id for VM, just re-use the
IB chunk id and add a new flags for VM mode. Also define additional
dwords for the flags chunk id to define the what ring we want to use
(gfx, compute, uvd, etc.) and the priority.
v19:
- fix cs fini in weird case of no ib
- semi working flush fix for ni
- rebase on top of sa allocator changes
v20: agd5f:
- further CS ioctl cleanups from Christian's comments
v21: agd5f:
- integrate CS checker improvements
v22: agd5f:
- final cleanups for release, only allow VM CS on cayman
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Somewhat specializaed sub-allocator designed to perform sub-allocation
for command buffer not only for current cs ioctl but for future command
submission ioctl as well. Patch also convert current ib pool to use
the sub allocator. Idea is that ib poll buffer can be share with other
command buffer submission not having 64K granularity.
v2 Harmonize pool handling and add suspend/resume callback to pin/unpin
sa bo (tested on rv280, rv370, r420, rv515, rv610, rv710, redwood, cayman,
rs480, rs690, rs880)
v3 Simplify allocator
v4 Fix radeon_ib_get error path to properly free fence
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Add a start fence driver helper function which will be call
once for each ring and will compute cpu/gpu addr for fence
depending on wether to use wb buffer or scratch reg.
This patch replace initialize fence driver separately which
was broken in regard of GPU lockup. The fence list for created,
emited, signaled must be initialize once and only from the
asic init callback not from the startup call back which is
call from the gpu reset.
v2: With this in place we no longer need to know the number of
rings in fence_driver_init, also writing to the scratch reg
before knowing its offset is a bad idea.
v3: rebase on top of change to previous patch in the serie
Signed-off-by: Christian König <deathsimple@vodafone.de>
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
That naming seems to make more sense, since we not
only want to run PM4 rings with it.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Replace cp, cp1 and cp2 members with just an array
of radeon_cp structs.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Give all asic and radeon_ring_* functions a
radeon_cp parameter, so they know the ring to work with.
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
For supporting multiple CP ring buffers, async DMA
engines and UVD. We still need a way to synchronize
between engines.
v2 initialize unused fence driver ring to avoid issue in
suspend/unload
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This adds a new optional chunk to the CS ioctl that specifies optional flags
to the CS parser. Why this is useful is explained below. Note that some regs
no longer need the NOP relocation packet if this feature is enabled.
Tested on r300g and r600g with this flag disabled and enabled.
Assume there are two contexts sharing the same mipmapped tiled texture.
One context wants to render into the first mipmap and the other one
wants to render into the last mipmap. As you probably know, the hardware
has a MACRO_SWITCH feature, which turns off macro tiling for small mipmaps,
but that only applies to samplers.
(at least on r300-r500, though later hardware likely behaves the same)
So we want to just re-set the tiling flags before rendering (writing
packets), right? ... No. The contexts run in parallel, so they may
set the tiling flags simultaneously and then fire their command streams
also simultaneously. The last one setting the flags wins, the other one
loses.
Another problem is when one context wants to render into the first and
the last mipmap in one CS. Impossible. It must flush before changing
tiling flags and do the rendering into the smaller mipmaps in another CS.
Yet another problem is that writing copy_blit in userspace would be a mess
involving re-setting tiling flags to please the kernel, and causing races
with other contexts at the same time.
The only way out of this is to send tiling flags with each CS, ideally
with each relocation. But we already do that through the registers.
So let's just use what we have in the registers.
Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
After GPU lockup VRAM gart table is unpinned and thus its pointer
becomes unvalid. This patch move the unpin code to a common helper
function and set pointer to NULL so that page update code can check
if it should update GPU page table or not. That way bo still bound
to GART can be unbound (pci_unmap_page for all there page) properly
while there is no need to update the GPU page table.
V2 move the test for null gart out of the loop, small optimization
Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This was previously done for r300 only. Use %016llX instead of %08X for
printing the table address.
Also fix typos in gart warning messages.
Signed-off-by: Tormod Volden <debian.tormod@gmail.com>
Reviewed-by: Michel Dänzer <michel@daenzer.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (47 commits)
doc: CONFIG_UNEVICTABLE_LRU doesn't exist anymore
Update cpuset info & webiste for cgroups
dcdbas: force SMI to happen when expected
arch/arm/Kconfig: remove one to many l's in the word.
asm-generic/user.h: Fix spelling in comment
drm: fix printk typo 'sracth'
Remove one to many n's in a word
Documentation/filesystems/romfs.txt: fixing link to genromfs
drivers:scsi Change printk typo initate -> initiate
serial, pch uart: Remove duplicate inclusion of linux/pci.h header
fs/eventpoll.c: fix spelling
mm: Fix out-of-date comments which refers non-existent functions
drm: Fix printk typo 'failled'
coh901318.c: Change initate to initiate.
mbox-db5500.c Change initate to initiate.
edac: correct i82975x error-info reported
edac: correct i82975x mci initialisation
edac: correct commented info
fs: update comments to point correct document
target: remove duplicate include of target/target_core_device.h from drivers/target/target_core_hba.c
...
Trivial conflict in fs/eventpoll.c (spelling vs addition)
This is an important security fix because we allowed arbitrary values
to be passed to AARESOLVE_OFFSET. This also puts the right buffer address
in the register.
Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The colorbuffer, zbuffer, and texture states are checked only once when
they get changed. This improves performance in the apps which emit
lots of draw packets and few state changes.
This drops performance in glxgears by a 1% or so, but glxgears is not
a benchmark we care about.
The time spent in the kernel when running Torcs dropped from 33% to 23%
and the frame rate is higher, which is a good thing.
r600 might need something like this as well.
Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
First, we were calling mc_stop() at the top of the function
which turns off all MC (memory controller) clients,
then checking if the GPU is idle. If it was idle we
returned without re-enabling the MC clients which would
lead to a blank screen, etc. This patch checks if the
GPU is idle before calling mc_stop().
Second, if the reset failed, we were returning without
re-enabling the MC clients. This patch re-enables
the MC clients before returning regardless of whether
the reset was successful or not.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The CMASK RAM is for colorbuffer compression (used in conjunction
with MSAA). Only one user (filp) can access it.
The CMASK RAM access is managed in the same way as Hyper-Z, but there is
a separate ioctl, because an app that uses MSAA does not necessarily
have to use zbuffering.
Signed-off-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
When writeback is enabled, the GPU shadows writes to certain
registers into a buffer in memory. The driver can then read
the values from the shadow rather than reading back from the
register across the bus. Writeback can be disabled by setting
the no_wb module param to 1.
On r6xx/r7xx/evergreen, the following registers are shadowed:
- CP scratch registers
- CP read pointer
- IH write pointer
On r1xx-rr5xx, the following registers are shadowed:
- CP scratch registers
- CP read pointer
v2:
- Combine wb patches for r6xx-evergreen and r1xx-r5xx
- Writeback is disabled on AGP boards since it tends to be
unreliable on AGP using the gart.
- Check radeon_wb_init return values properly.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This interface allows userspace to request hyperz support, it probably
needs more locking, and really reporting that you can have hyperz is racy
since someone else might get it before you do.
v2: modify so we pass 0 valued packets to let DDX/r300c keep working.
also fixed incorrect 0x4f1c reference.
v3: fixup zb_bw_cntl so older drivers keep working
v4: add locking, fixup SC_HYPERZ_EN - patch stream to disable hiz
Signed-off-by: Dave Airlie <airlied@redhat.com>
On systems using kexec, the new kernel is booted straight from the old kernel, without any warning to the graphics driver. So the GPU is basically left as-is in a running state, however the CPU side is completly reset.
Without stating the saneness of anyone using kexec on live systems, we should at least try not to crash the GPU. This patch resets 3 registers to 0 that could cause bad things to happen to the running system.
This allows kexec to work on a Power6/RN50 system.
Signed-off-by: Dave Airlie <airlied@redhat.com>
The asics in question have the following requirements with regard to
their gart setups:
1. The GART aperture size has to be in the form of 2^X bytes, where X is from 25 to 31
2. The GART aperture MC base has to be aligned to a boundary equal to the size of the
aperture.
3. The GART page table has to be aligned to the boundary equal to the size of the table.
4. The GART page table size is: table_entry_size * (aperture_size / page_size)
5. The GART page table has to be allocated in non-paged, non-cached, contiguous system
memory.
This patch takes care 2. The rest should already be handled properly.
This fixes a regression noticed by: Torsten Kaiser <just.for.lkml@googlemail.com>
Tested-by: Torsten Kaiser <just.for.lkml@googlemail.com>
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=28459
agd5f: apply to r1xx/r2xx as well.
Signed-off-by: Roland Scheidegger <sroland@vmware.com>
Cc: stable@kernel.org
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
