- 04 Dec, 2012 6 commits
-
-
Inki Dae authored
This patch adds userptr feautre for G2D module. The userptr means user space address allocated by malloc(). And the purpose of this feature is to make G2D's dma able to access the user space region. To user this feature, user should flag G2D_BUF_USRPTR to offset variable of struct drm_exynos_g2d_cmd and fill struct drm_exynos_g2d_userptr with user space address and size for it and then should set a pointer to drm_exynos_g2d_userptr object to data variable of struct drm_exynos_g2d_cmd. The last bit of offset variable is used to check if the cmdlist's buffer type is userptr or not. If userptr, the g2d driver gets user space address and size and then gets pages through get_user_pages(). (another case is counted as gem handle) Below is sample codes: static void set_cmd(struct drm_exynos_g2d_cmd *cmd, unsigned long offset, unsigned long data) { cmd->offset = offset; cmd->data = data; } static int solid_fill_test(int x, int y, unsigned long userptr) { struct drm_exynos_g2d_cmd cmd_gem[5]; struct drm_exynos_g2d_userptr g2d_userptr; unsigned int gem_nr = 0; ... g2d_userptr.userptr = userptr; g2d_userptr.size = x * y * 4; set_cmd(&cmd_gem[gem_nr++], DST_BASE_ADDR_REG | G2D_BUF_USERPTR, (unsigned long)&g2d_userptr); ... } int main(int argc, char **argv) { unsigned long addr; ... addr = malloc(x * y * 4); ... solid_fill_test(x, y, addr); ... } And next, the pages are mapped with iommu table and the device address is set to cmdlist so that G2D's dma can access it. As you may know, the pages from get_user_pages() are pinned. In other words, they CAN NOT be migrated and also swapped out. So the dma access would be safe. But the use of userptr feature has performance overhead so this patch also has memory pool to the userptr feature. Please, assume that user sends cmdlist filled with userptr and size every time to g2d driver, and the get_user_pages funcion will be called every time. The memory pool has maximum 64MB size and the userptr that user had ever sent, is holded in the memory pool. This meaning is that if the userptr from user is same as one in the memory pool, device address to the userptr in the memory pool is set to cmdlist. And last, the pages from get_user_pages() will be freed once user calls free() and the dma access is completed. Actually, get_user_pages() takes 2 reference counts if the user process has never accessed user region allocated by malloc(). Then, if the user calls free(), the page reference count becomes 1 and becomes 0 with put_page() call. And the reverse holds as well. This means how the pages backed are used by dma and freed. This patch is based on "drm/exynos: add iommu support for g2d", https://patchwork.kernel.org/patch/1629481/Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Prathyush K authored
The function dma_get_sgtable will allocate a sg table internally so it is not necessary to allocate a sg table before it. The unnecessary 'sg_alloc_table' call is removed. Signed-off-by: Prathyush K <prathyush.k@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Rahul Sharma authored
This patch fixes the problem of mapping contigous and non contigous dma buffers. Currently page struct is calculated from the buf->dma_addr which is not the physical address. It is replaced by buf->pages which points to the page struct of the first page of contigous memory chunk. This gives the correct page frame number for mapping. Non-contigous dma buffers are described using SG table and SG lists. Each valid SG List is pointing to a single page or group of pages which are physically contigous. Current implementation just maps the first page of each SG List and leave the other pages unmapped, leading to a crash. Given solution finds the page struct for the faulting page through parsing SG table and map it. Signed-off-by: Rahul Sharma <rahul.sharma@samsung.com> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Inki Dae authored
With iommu support, non-continuous buffer also is supported so this patch removes these checking from exynos_drm_gem_get/put_dma_addr funciton. This patch is based on the below patch set, "drm/exynos: add iommu support for -next". http://www.spinics.net/lists/dri-devel/msg29041.htmlSigned-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Inki Dae authored
Chagelog v2: removed unnecessary structure, struct g2d_gem_node. Chagelog v1: This patch adds iommu support for g2d driver. For this, it adds subdrv_probe/remove callback to enable or disable g2d iommu. And with this patch, in case of using g2d iommu, we can get or put device address to a gem handle from user through exynos_drm_gem_get/put_dma_addr(). Actually, these functions take a reference to a gem handle so that the gem object used by g2d dma is released properly. And runqueue_node has a pointer to drm_file object of current process to manage gem handles to owner. This patch is based on the below patch set, "drm/exynos: add iommu support for -next". http://www.spinics.net/lists/dri-devel/msg29041.htmlSigned-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Inki Dae authored
Changelog v2: move iommu support feature to mixer side. And below is Prathyush's comment. According to the new IOMMU framework for exynos sysmmus, the owner of the sysmmu-tv is mixer (which is the actual device that does DMA) and not hdmi. The mmu-master in sysmmu-tv node is set as below in exynos5250.dtsi sysmmu-tv { - mmu-master = <&mixer>; }; Changelog v1: The iommu will be enabled when hdmi sub driver is probed and will be disabled when removed. Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
- 29 Nov, 2012 17 commits
-
-
Inki Dae authored
The iommu will be enabled when fimd sub driver is probed and will be disabled when removed. Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Inki Dae authored
Changelog v4: - fix condition to drm_iommu_detach_device funtion. Changelog v3: - add dma_parms->max_segment_size setting of drm_device->dev. - use devm_kzalloc instead of kzalloc. Changelog v2: - fix iommu attach condition. . check archdata.dma_ops of drm device instead of subdrv device's one. - code clean to exynos_drm_iommu.c file. . remove '#ifdef CONFIG_ARM_DMA_USE_IOMMU' from exynos_drm_iommu.c and add it to driver/gpu/drm/exynos/Kconfig. Changelog v1: This patch adds iommu support for exynos drm framework with dma mapping api. In this patch, we used dma mapping api to allocate physical memory and maps it with iommu table and removed some existing codes and added new some codes for iommu support. GEM allocation requires one device object to use dma mapping api so this patch uses one iommu mapping for all sub drivers. In other words, all sub drivers have same iommu mapping. Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Marek Szyprowski authored
This patch adds support for DMA_ATTR_FORCE_CONTIGUOUS attribute for dma_alloc_attrs() in IOMMU-aware implementation. For allocating physically contiguous buffers Contiguous Memory Allocator is used. Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
-
Marek Szyprowski authored
This patch adds DMA_ATTR_FORCE_CONTIGUOUS attribute to the DMA-mapping subsystem. By default DMA-mapping subsystem is allowed to assemble the buffer allocated by dma_alloc_attrs() function from individual pages if it can be mapped as contiguous chunk into device dma address space. By specifing this attribute the allocated buffer is forced to be contiguous also in physical memory. Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com>
-
Imre Deak authored
Currently the only users of drm_vblank_off() are i915 and gma500, neither of which holds the event_lock when calling this function. Fix this by holding the event_lock while traversing the list. Signed-off-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Imre Deak authored
Currently the exynos driver calls drm_vblank_off() with the event_lock held, while drm_vblank_off() will lock vbl_time and vblank_time_lock. This lock dependency chain conflicts with the one in drm_handle_vblank() where we first lock vblank_time_lock and then the event_lock. Fix this by removing the above drm_vblank_off() calls which are in fact never executed: drm_dev->vblank_disable_allowed is only ever non-zero during driver init, until it's set in {fimd,vidi}_subdrv_probe. Both the driver init and open code is protected by drm_global_mutex, so the earliest page flip ioctl can happen only after vblank_disable_allowed is set to 1. Thus {fimd,vidi}_finish_pageflip - with pending flip events - will always get called with vblank_disable_allowed being 1. Signed-off-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Imre Deak authored
It's guaranteed that for each event on pageflip_event_list we have called drm_vblank_get() - see exynos_drm_crtc_page_flip() - so checking for this is redundant. Also we need to call drm_vblank_put() for each event on the list, not only once, otherwise we'd leak vblank references if there are multiple events on the list. Signed-off-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Imre Deak authored
Signed-off-by: Imre Deak <imre.deak@intel.com> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Sachin Kamat authored
Check overlay_ops is not NULL as checked in the previous 'if' condition. Fixes the following smatch error: drivers/gpu/drm/exynos/exynos_drm_encoder.c:509 exynos_drm_encoder_plane_disable() error: we previously assumed 'overlay_ops' could be null (see line 499) Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Sachin Kamat authored
Fixes the following sparse warnings: drivers/gpu/drm/exynos/exynos_drm_fimd.c:65:25: warning: symbol 'exynos4_fimd_driver_data' was not declared. Should it be static? drivers/gpu/drm/exynos/exynos_drm_fimd.c:69:25: warning: symbol 'exynos5_fimd_driver_data' was not declared. Should it be static? Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Inki Dae authored
Chagelog v2: Move encoder's dpms updating into exynos_drm_encoder_commit function because when crtc's dpms is updated, encoder's dpms is updated also. This would induce the issue that encoder isn't disabled after crtc is disabled. Changelog v1: This patch fixes a issue that overlay data aren't applied to real hardware when dpms off goes to on after setcrtc was requested like below, dpms off -> setcrtc -> dpms off -> dpms on For this, it makes encoder's dpms to be updated when setcrtc is requested. Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Inki Dae authored
plane->fb will be set to new fb after update_plane callback is called by drm_mode_set_plane() Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
Inki Dae authored
With iommu, buffer->dma_addr has device addres so this patch fixes for physical address to be set to fix.smem_start always. Signed-off-by: Inki Dae <inki.dae@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
-
git://people.freedesktop.org/~danvet/drm-intelDave Airlie authored
Daniel writes: Besides the big item of lifting the "preliminary hw support" tag from the Haswell code, just small bits&pieces all over: - Leftover Haswell patches and some fixes from Paulo - LyncPoint PCH support (for hsw) - OOM handling improvements from Chris Wilson - connector property and send_vblank_event refactorings from Rob Clark - random pile of small fixes Note that the send_vblank refactorings will cause some locking WARNs to show up. Imre has fixed that up, but since all the driver changes outside of the drm core have been for exonys, those four patches are merged through the exonys-next tree. Meh, I've forgotten to cherry-pick an important fix from Ben for a regression in the 3.8 gen6+ gtt code. New pull request below. While I'm at it, the hdmi VIC patch for the drm edid code is still in my queue, I'll send you that in the next 3.8-fixes pull. * 'for-airlied' of git://people.freedesktop.org/~danvet/drm-intel: (33 commits) drm/i915: Fix pte updates in ggtt clear range drm/i915: promote Haswell to full support drm/i915: Report the origin of the LVDS fixed panel mode drm/i915: LVDS fallback to fixed-mode if EDID not present drm/i915/sdvo: kfree the intel_sdvo_connector, not drm_connector, on destroy drm/i915: drm_connector_property -> drm_object_property drm/i915: use drm_send_vblank_event() helper drm/i915: Use pci_resource functions for BARs. drm/i915: Borrow our struct_mutex for the direct reclaim drm/i915: Defer assignment of obj->gtt_space until after all possible mallocs drm/i915: Apply the IBX transcoder A w/a for HDMI to SDVO as well drm/i915: implement WaMbcDriverBootEnable on Haswell drm/i915: fix intel_ddi_get_cdclk_freq for ULT machines drm/i915: make the panel fitter work on pipes B and C on Haswell drm/i915: make the panel fitter work on pipes B and C on IVB drm/i915: don't intel_crt_init if DDI A has 4 lanes drm/i915: make DP work on LPT-LP machines drm/i915: fix false positive "Unclaimed write" messages drm/i915: use cpu/pch transcoder on intel_enable_pipe drm/i915: don't limit Haswell CRT encoder to pipe A ...
-
Ben Widawsky authored
This bug was introduced by me: commit e76e9aeb Author: Ben Widawsky <ben@bwidawsk.net> Date: Sun Nov 4 09:21:27 2012 -0800 drm/i915: Stop using AGP layer for GEN6+ The existing code uses memset_io which follows memset semantics in only guaranteeing a write of individual bytes. Since a PTE entry is 4 bytes, this can only be correct if the scratch page address is 0. This caused unsightly errors when we clear the range at load time, though I'm not really sure what the heck is referencing that memory anyway. I caught this is because I believe we have some other bug where the display is doing reads of memory we feel should be cleared (or we are relying on scratch pages to be a specific value). Signed-off-by: Ben Widawsky <ben@bwidawsk.net> Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
-
Thierry Reding authored
Add myself as the maintainer of the NVIDIA Tegra DRM driver. Signed-off-by: Thierry Reding <thierry.reding@avionic-design.de> Acked-by: Stephen Warren <swarren@nvidia.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
Jingoo Han authored
Fixed build warning as below: drivers/gpu/drm/drm_pci.c: In function 'drm_pcie_get_speed_cap_mask': drivers/gpu/drm/drm_pci.c:496:9: warning: 'lnkcap' may be used uninitialized in this function [-Wuninitialized] drivers/gpu/drm/drm_pci.c:497:10: warning: 'lnkcap2' may be used uninitialized in this function [-Wuninitialized] Signed-off-by: Jingoo Han <jg1.han@samsung.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
-
- 28 Nov, 2012 17 commits
-
-
Sachin Kamat authored
core/device.h was included twice. Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Sachin Kamat authored
nouveau_ttm.h was included twice. Signed-off-by: Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Ben Skeggs authored
Fixes a PCE0 page fault noticed on NVD9 during module unload. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Daniel J Blueman authored
On 3.7-rc6, add missing newline to to prevent the following kernel log line getting appended to the current one after switching the integrated GPU and suspending the discrete GPU. Signed-off-by: Daniel J Blueman <daniel@quora.org> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Marcin Slusarz authored
These objects leak VRAM - but only on module unload. Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Marcin Slusarz authored
Fixes GART leak (as accounted by nouveau_drm.gem.gart_available). Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Marcin Slusarz authored
Copy/pasted from nouveau_gem_new. Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Marcin Slusarz authored
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Martin Peres authored
It fixes a bug that would have been introduced when adding more sudevs/engines. Signed-off-by: Martin Peres <martin.peres@labri.fr> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Ben Skeggs authored
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Ben Skeggs authored
NVIDIA also appear to use the same class on Fermi/Kepler for PPP. Will allow use of the engine if firmware (nvXX_fuc086) provided. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Maarten Lankhorst authored
Will allow use of the engine if firmware (nvXX_fuc086) provided. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Maarten Lankhorst authored
Will allow use of the engine if firmware (nvXX_fuc085) provided. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Maarten Lankhorst authored
Will allow use of the engine if firmware (nvXX_fuc084) provided. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Ben Skeggs authored
Later chipsets use falcon anyway, and I can't currently see a good need for a shared base class. PPP will get the same treatment once Maarten's patches are merged. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Ben Skeggs authored
No support for the class yet, but will be pulled in with Maarten's Fermi vdec patches. The Kepler PPP class is identical. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-
Ben Skeggs authored
Will allow use of the engine if firmware (nvXX_fuc085) provided. Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
-