summaryrefslogtreecommitdiff
path: root/nouveau
AgeCommit message (Collapse)Author
2013-03-26nouveau: add a way to override single pushbuffer memory limitsMarcin Slusarz
Currently single pushbuffer can take up to 80% of VRAM and 80% of GART. As this value seems to be arbitrary (and user may need to set it differently) this patch adds support for 2 environment variables: NOUVEAU_LIBDRM_VRAM_LIMIT_PERCENT (default 80) NOUVEAU_LIBDRM_GART_LIMIT_PERCENT (default 80) which will let users override pushbuffer VRAM/GART limits. Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
2013-03-26nouveau: return error from pushbuf_validateMarcin Slusarz
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
2013-01-16nouveau: use @PACKAGE_VERSION@ in libdrm_nouveau.pcMaarten Lankhorst
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
2013-01-16nouveau: expose channel engine selection on kepler chipsetsBen Skeggs
v2: Take Maarten Lankhorst's suggestion of nesting the struct to prevent sizeof() issues due to padding on older revisions. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
2013-01-14libdrm_nouveau.pc: don't include I${includedir}/nouveauMarcin Slusarz
Nouveau headers are installed in I${includedir}/libdrm.
2013-01-07nouveau: disallow pushbuf BOs in multiple memory typesBen Skeggs
Under certain circumstances it's possible for libdrm to decide to move a GART|VRAM pushbuf to be VRAM-only. This causes the kernel to reject the command submission on GF8 and up, due to a stricter policy where buffers are only allowed to move to memory types that were specified at creation time. The simplest fix for this is to force the creation-time memory type for the lifetime of the push buffer. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2012-07-19libdrm/nouveau: add prime handle->bo and bo->handle support.Dave Airlie
This adds prime support to nouveau libdrm. Signed-off-by: Dave Airlie <airlied@redhat.com>
2012-06-09nouveau: silence some remaining valgrind warningsMarcin Slusarz
Valgrind can't understand some of the fields passed to ioctls are overwritten by kernel, so we need to initialize them. Almost all of our ioctl wrappers already do it and the cost of remaining 3 is very small. Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
2012-05-02nouveau: fix channel closingMarcin Slusarz
Restore code lost in libdrm_nouveau rewrite. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2012-04-24nouveau: expose notifier handle on nvc0 as wellChristoph Bumiller
Signed-off-by: Christoph Bumiller <e0425955@student.tuwien.ac.at> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2012-04-24nouveau: remove unnecessary EAGAIN loopsMarcin Slusarz
drmCommandWrite / drmCommandWriteRead already loop on EAGAIN. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2012-04-17nouveau: init nvc0 channel alloc req structure fullyBen Skeggs
Kernel rejects ~0 handles, even though they're not used on NVC0. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2012-04-14nouveau: pull in major libdrm rewriteBen Skeggs
Redesigned primarily to allow us to better take advantage of BO's having fixed GPU virtual addresses on GeForce 8 and up, and to reduce the overhead of handling relocations on earlier chipsets. Signed-off-by: Ben Skeggs <bskeggs@redhat.com> Signed-off-by: Christoph Bumiller <e0425955@student.tuwien.ac.at>
2011-11-27nouveau: Mark nouveau subchannel unbound nouveau_grobj_freeMaarten Lankhorst
Valgrind throws warns about a user-after-free if you try to bind a new subchannel after the old one in that slot was freed, so remove it from the channel list. Signed-off-by: Maarten Lankhorst <m.b.lankhorst@gmail.com>
2011-10-19nouveau: free in error path if drmAvailable fails.Dave Airlie
This was reported in coverity. Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-09-18nouveau: assert argument cannot have side effectsMarcin Slusarz
... because argument is evaluated only if NDEBUG is not defined
2010-12-21nouveau: fix up reloc_emit() to accept NULL target bufferBen Skeggs
The nvc0 gallium drivers passes NULL here to indicate to the memory manager that a buffer is being used, but without creating an actual reloc. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-12-21nouveau: split pushbuf macros specific to nv04-nv50 out, and add nvc0Ben Skeggs
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-12-21nouveau: nvc0 drm has no concept of "notifier block"Ben Skeggs
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-11-22nouveau: Add implicit pushbuf flush before gpuobj destruction.Francisco Jerez
It makes sure that GPU object destruction is executed in order with respect to the previous FIFO commands. Signed-off-by: Francisco Jerez <currojerez@riseup.net> Acked-by: Ben Skeggs <bskeggs@redhat.com>
2010-10-31nouveau: Avoid unnecessary call to CPU_FINI.Francisco Jerez
nouveau_bo_unmap called the CPU_FINI IOCTL even if it was a NOSYNC mapping. It caused no harmful effects (actually CPU_FINI is a no-op on recent enough kernels) besides the precious CPU cycles being wasted. Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-10-12nouveau: Let the user choose the push buffer size.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net> Acked-by: Ben Skeggs <bskeggs@redhat.com>
2010-10-12nouveau: Define buffer object usage flags.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net> Acked-by: Ben Skeggs <bskeggs@redhat.com>
2010-09-21nouveau: fix drm version check some moreBen Skeggs
... and make a mental note to not push commits before having coffee Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-09-20nouveau: fix thinko in drm version checkBen Skeggs
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-08-18nouveau: accept both 0.0.16 and 1.x.xBen Skeggs
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-04-09nouveau: stop shipping nouveau_class.hBen Skeggs
The motivation behind this is that by shipping it here, it's essentially an API which causes issues while bisecting across changes to the header files.
2010-03-19nouveau: fix annoying compiler warningBen Skeggs
2010-03-19nouveau: fix segfault in nouveau_bo_new_tile() failure pathBen Skeggs
2010-03-19nouveau: remove unused field from nouveau_boBen Skeggs
2010-03-18nouveau: Regenerate nouveau_class.h.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-03-17Install headers to $(includedir)/libdrmJulien Cristau
Avoids conflicts with kernel headers. Signed-off-by: Julien Cristau <jcristau@debian.org> Reviewed-by: Rémi Cardona <remi@gentoo.org> Signed-off-by: Eric Anholt <eric@anholt.net>
2010-03-17libdrm_nouveau requires libdrmJulien Cristau
nouveau_drmif.h includes xf86drm.h. Signed-off-by: Julien Cristau <jcristau@debian.org> Signed-off-by: Eric Anholt <eric@anholt.net>
2010-03-13nouveau: Fix up the stride of NV20TCL_LIGHT_BACK_*.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-03-07nouveau: Small lighting related addition to nouveau_class.h.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-03-06nouveau: Update nouveau_class.h.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-03-02nouveau: make sure initial kalloc for user bo ends up in the right placeMaarten Maathuis
- Currently reloc'ing a user bo to gart will first cause an allocation in vram, which is then written to by cpu, then the bo gets moved to gart. Acked-by: Francisco Jerez <currojerez@riseup.net> Signed-off-by: Maarten Maathuis <madman2003@gmail.com>
2010-02-25nouveau: Update nouveau_class.h.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-02-20nv30: update for 8 texture unitsPatrice Mandin
2010-02-20nv30: update for front/back stencil inversionPatrice Mandin
Signed-off-by: Patrice Mandin <patmandin@gmail.com>
2010-02-16nouveau: interface changes for 0.0.16 DRMLuca Barbieri
This commit encompasses the changes necessary to run on top of the 0.0.16 nouveau interface, additional APIs to support the new features of the interface, as well as code from Luca Barbieri to improve the pushbuf interface, which just happens to break nouveau's libdrm ABI so was delayed until now. API changes as a result of 0.0.16 DRM interface: 1. No more bo_pin()/bo_unpin(), these were only there for UMS and we no longer support it. 2. Any random nouveau_bo can be submitted to the GPU as a push buffer. 3. Relocations can be applied on any nouveau_bo This patch changes the pushbuffer ABI to: 1. No longer use/expose nouveau_pushbuffer. Everything is directly in nouveau_channel. This saves the extra "pushbuf" pointer dereference. 2. Use cur/end pointers instead of tracking the remaining size. Pushing data now only needs to alter cur and not both cur and remaining. The goal is to make the *_RING macros faster and make the interface simpler and cleaner in the process. The *_RING APIs are unchanged, but those are inlined and the ABI is changed. Also, anything accessing pushbuf->remaining instead of using AVAIL_RING will need to be fixed.
2010-02-01nouveau: Regenerate nouveau_class.h.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-01-31nouveau: add nouveau_resource_destroyMarcin Slusarz
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2010-01-21nouveau: fail channel creation if pushbuf init failsBen Skeggs
2010-01-15nouveau: disable flush_notify on channel_freeMarcin Slusarz
We don't want do call flush_notify when we will FIRE the RING a couple of lines later, because grobj bound to this channel might be already freed.
2010-01-15nouveau: fix memory leak in nouveau_channel_freeMarcin Slusarz
2010-01-08nouveau: Update nouveau_class.h.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
2009-12-30nouveau: add BEGIN_RING_NIMarcin Slusarz
2009-12-28nouveau: Unreference pushbuf objects on channel destruction.Younes Manton
- unreference pushbuf objects on channel destruction Based on Krzysztof Smiechowicz's patch.
2009-12-28nouveau: Update some object definitions from renouveau.xml.Francisco Jerez
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-26 18:55:34 +0000