From 510fe10b6180137773dce9032b51bb82ff946c2d Mon Sep 17 00:00:00 2001 From: Zhao Yan Date: Tue, 19 Jun 2018 15:44:11 +0800 Subject: drm/i915/gvt: fix a bug of partially write ggtt enties when guest writes ggtt entries, it could write 8 bytes a time if gtt_entry_size is 8. But, qemu could split the 8 bytes into 2 consecutive 4-byte writes. If each 4-byte partial write could trigger a host ggtt write, it is very possible that a wrong combination is written to the host ggtt. E.g. the higher 4 bytes is the old value, but the lower 4 bytes is the new value, and this 8-byte combination is wrong but written to the ggtt, thus causing bugs. To handle this condition, we just record the first 4-byte write, then wait until the second 4-byte write comes and write the combined 64-bit data to host ggtt table. To save memory space and to spot partial write as early as possible, we don't keep this information for every ggtt index. Instread, we just record the last ggtt write position, and assume the two 4-byte writes come in consecutively for each vgpu. This assumption is right based on the characteristic of ggtt entry which stores memory address. When gtt_entry_size is 8, the guest memory physical address should be 64 bits, so any sane guest driver should write 8-byte long data at a time, so 2 consecutive 4-byte writes at the same ggtt index should be trapped in gvt. v2: when incomplete ggtt entry write is located, e.g. 1. guest only writes 4 bytes at a ggtt offset and no long writes the rest 4 bytes. 2. guest writes 4 bytes of a ggtt offset, then write at other ggtt offsets, then return back to write the left 4 bytes of the first ggtt offset. add error handling logic to remap host entry to scratch page, and mark guest virtual ggtt entry as not present. (zhenyu wang) Signed-off-by: Zhao Yan Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/gtt.c | 58 ++++++++++++++++++++++++++++++++++++++++++ drivers/gpu/drm/i915/gvt/gtt.h | 2 ++ 2 files changed, 60 insertions(+) (limited to 'drivers/gpu/drm') diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c index 23296547da95..4efec8fa6c1d 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.c +++ b/drivers/gpu/drm/i915/gvt/gtt.c @@ -1592,6 +1592,7 @@ static struct intel_vgpu_mm *intel_vgpu_create_ggtt_mm(struct intel_vgpu *vgpu) vgpu_free_mm(mm); return ERR_PTR(-ENOMEM); } + mm->ggtt_mm.last_partial_off = -1UL; return mm; } @@ -1616,6 +1617,7 @@ void _intel_vgpu_mm_release(struct kref *mm_ref) invalidate_ppgtt_mm(mm); } else { vfree(mm->ggtt_mm.virtual_ggtt); + mm->ggtt_mm.last_partial_off = -1UL; } vgpu_free_mm(mm); @@ -1868,6 +1870,62 @@ static int emulate_ggtt_mmio_write(struct intel_vgpu *vgpu, unsigned int off, memcpy((void *)&e.val64 + (off & (info->gtt_entry_size - 1)), p_data, bytes); + /* If ggtt entry size is 8 bytes, and it's split into two 4 bytes + * write, we assume the two 4 bytes writes are consecutive. + * Otherwise, we abort and report error + */ + if (bytes < info->gtt_entry_size) { + if (ggtt_mm->ggtt_mm.last_partial_off == -1UL) { + /* the first partial part*/ + ggtt_mm->ggtt_mm.last_partial_off = off; + ggtt_mm->ggtt_mm.last_partial_data = e.val64; + return 0; + } else if ((g_gtt_index == + (ggtt_mm->ggtt_mm.last_partial_off >> + info->gtt_entry_size_shift)) && + (off != ggtt_mm->ggtt_mm.last_partial_off)) { + /* the second partial part */ + + int last_off = ggtt_mm->ggtt_mm.last_partial_off & + (info->gtt_entry_size - 1); + + memcpy((void *)&e.val64 + last_off, + (void *)&ggtt_mm->ggtt_mm.last_partial_data + + last_off, bytes); + + ggtt_mm->ggtt_mm.last_partial_off = -1UL; + } else { + int last_offset; + + gvt_vgpu_err("failed to populate guest ggtt entry: abnormal ggtt entry write sequence, last_partial_off=%lx, offset=%x, bytes=%d, ggtt entry size=%d\n", + ggtt_mm->ggtt_mm.last_partial_off, off, + bytes, info->gtt_entry_size); + + /* set host ggtt entry to scratch page and clear + * virtual ggtt entry as not present for last + * partially write offset + */ + last_offset = ggtt_mm->ggtt_mm.last_partial_off & + (~(info->gtt_entry_size - 1)); + + ggtt_get_host_entry(ggtt_mm, &m, last_offset); + ggtt_invalidate_pte(vgpu, &m); + ops->set_pfn(&m, gvt->gtt.scratch_mfn); + ops->clear_present(&m); + ggtt_set_host_entry(ggtt_mm, &m, last_offset); + ggtt_invalidate(gvt->dev_priv); + + ggtt_get_guest_entry(ggtt_mm, &e, last_offset); + ops->clear_present(&e); + ggtt_set_guest_entry(ggtt_mm, &e, last_offset); + + ggtt_mm->ggtt_mm.last_partial_off = off; + ggtt_mm->ggtt_mm.last_partial_data = e.val64; + + return 0; + } + } + if (ops->test_present(&e)) { gfn = ops->get_pfn(&e); m = e; diff --git a/drivers/gpu/drm/i915/gvt/gtt.h b/drivers/gpu/drm/i915/gvt/gtt.h index 3792f2b7f4ff..97e62647418a 100644 --- a/drivers/gpu/drm/i915/gvt/gtt.h +++ b/drivers/gpu/drm/i915/gvt/gtt.h @@ -150,6 +150,8 @@ struct intel_vgpu_mm { } ppgtt_mm; struct { void *virtual_ggtt; + unsigned long last_partial_off; + u64 last_partial_data; } ggtt_mm; }; }; -- cgit v1.2.3 From a4cae23cc05ccec749c2fc70fa9d8cda7c582319 Mon Sep 17 00:00:00 2001 From: Xiaolin Zhang Date: Thu, 21 Jun 2018 14:33:43 +0800 Subject: drm/i915/gvt: changed DDI mode emulation type changed gvt display transcode DDI mode from DP_SST to DVI to address below calltrace issue during guest booting up which is caused by zero dotclock initial value with DP_SST mode. transcode DVI mode emulation also align with native with DP connection. [drm:drm_calc_timestamping_constants] ERROR crtc 41: Can't calculate constants, dotclock = 0! WARNING: at drivers/gpu/drm/drm_vblank.c:620 drm_calc_vbltimestamp_from_scanoutpos Call Trace: ? drm_calc_timestamping_constants+0x144/0x150 [drm] drm_get_last_vbltimestamp+0x54/0x90 [drm] drm_reset_vblank_timestamp+0x59/0xd0 [drm] drm_crtc_vblank_on+0x7b/0xd0 [drm] intel_modeset_setup_hw_state+0xb67/0xfd0 [i915] ? gen2_read32+0x110/0x110 [i915] ? drm_modeset_lock+0x30/0xa0 [drm] intel_modeset_init+0x794/0x19d0 [i915] ? intel_setup_gmbus+0x232/0x2e0 [i915] i915_driver_load+0xb4a/0xf40 [i915] Signed-off-by: Xiaolin Zhang Signed-off-by: Zhenyu Wang --- drivers/gpu/drm/i915/gvt/display.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'drivers/gpu/drm') diff --git a/drivers/gpu/drm/i915/gvt/display.c b/drivers/gpu/drm/i915/gvt/display.c index 6d8180e8d1e2..4b072ade8c38 100644 --- a/drivers/gpu/drm/i915/gvt/display.c +++ b/drivers/gpu/drm/i915/gvt/display.c @@ -196,7 +196,7 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) ~(TRANS_DDI_BPC_MASK | TRANS_DDI_MODE_SELECT_MASK | TRANS_DDI_PORT_MASK); vgpu_vreg_t(vgpu, TRANS_DDI_FUNC_CTL(TRANSCODER_A)) |= - (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | + (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DVI | (PORT_B << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); if (IS_BROADWELL(dev_priv)) { @@ -216,7 +216,7 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) ~(TRANS_DDI_BPC_MASK | TRANS_DDI_MODE_SELECT_MASK | TRANS_DDI_PORT_MASK); vgpu_vreg_t(vgpu, TRANS_DDI_FUNC_CTL(TRANSCODER_A)) |= - (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | + (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DVI | (PORT_C << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); if (IS_BROADWELL(dev_priv)) { @@ -236,7 +236,7 @@ static void emulate_monitor_status_change(struct intel_vgpu *vgpu) ~(TRANS_DDI_BPC_MASK | TRANS_DDI_MODE_SELECT_MASK | TRANS_DDI_PORT_MASK); vgpu_vreg_t(vgpu, TRANS_DDI_FUNC_CTL(TRANSCODER_A)) |= - (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DP_SST | + (TRANS_DDI_BPC_8 | TRANS_DDI_MODE_SELECT_DVI | (PORT_D << TRANS_DDI_PORT_SHIFT) | TRANS_DDI_FUNC_ENABLE); if (IS_BROADWELL(dev_priv)) { -- cgit v1.2.3 From d403397c7c0821704a2f4da2694b46e423791261 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Sat, 30 Jun 2018 10:05:09 +0100 Subject: drm/i915: Try GGTT mmapping whole object as partial If the whole object is already pinned by HW for use as scanout, we will fail to move it to the mappable region and so must resort to using a partial VMA covering the whole object. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=104513 Fixes: aa136d9d72c2 ("drm/i915: Convert partial ggtt vma to full ggtt if it spans the entire object") Signed-off-by: Chris Wilson Cc: Joonas Lahtinen Cc: Matthew Auld Reviewed-by: Matthew Auld Link: https://patchwork.freedesktop.org/patch/msgid/20180630090509.469-1-chris@chris-wilson.co.uk (cherry picked from commit 7e7367d3bc6cf27dd7e007e7897fcebfeff1ee8b) Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_gem.c | 28 +++++++++++++++++----------- drivers/gpu/drm/i915/i915_vma.c | 2 +- 2 files changed, 18 insertions(+), 12 deletions(-) (limited to 'drivers/gpu/drm') diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index d44ad7bc1e94..17c5097721e8 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -2002,7 +2002,6 @@ int i915_gem_fault(struct vm_fault *vmf) bool write = !!(vmf->flags & FAULT_FLAG_WRITE); struct i915_vma *vma; pgoff_t page_offset; - unsigned int flags; int ret; /* We don't use vmf->pgoff since that has the fake offset */ @@ -2038,27 +2037,34 @@ int i915_gem_fault(struct vm_fault *vmf) goto err_unlock; } - /* If the object is smaller than a couple of partial vma, it is - * not worth only creating a single partial vma - we may as well - * clear enough space for the full object. - */ - flags = PIN_MAPPABLE; - if (obj->base.size > 2 * MIN_CHUNK_PAGES << PAGE_SHIFT) - flags |= PIN_NONBLOCK | PIN_NONFAULT; /* Now pin it into the GTT as needed */ - vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, flags); + vma = i915_gem_object_ggtt_pin(obj, NULL, 0, 0, + PIN_MAPPABLE | + PIN_NONBLOCK | + PIN_NONFAULT); if (IS_ERR(vma)) { /* Use a partial view if it is bigger than available space */ struct i915_ggtt_view view = compute_partial_view(obj, page_offset, MIN_CHUNK_PAGES); + unsigned int flags; - /* Userspace is now writing through an untracked VMA, abandon + flags = PIN_MAPPABLE; + if (view.type == I915_GGTT_VIEW_NORMAL) + flags |= PIN_NONBLOCK; /* avoid warnings for pinned */ + + /* + * Userspace is now writing through an untracked VMA, abandon * all hope that the hardware is able to track future writes. */ obj->frontbuffer_ggtt_origin = ORIGIN_CPU; - vma = i915_gem_object_ggtt_pin(obj, &view, 0, 0, PIN_MAPPABLE); + vma = i915_gem_object_ggtt_pin(obj, &view, 0, 0, flags); + if (IS_ERR(vma) && !view.type) { + flags = PIN_MAPPABLE; + view.type = I915_GGTT_VIEW_PARTIAL; + vma = i915_gem_object_ggtt_pin(obj, &view, 0, 0, flags); + } } if (IS_ERR(vma)) { ret = PTR_ERR(vma); diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c index 9324d476e0a7..0531c01c3604 100644 --- a/drivers/gpu/drm/i915/i915_vma.c +++ b/drivers/gpu/drm/i915/i915_vma.c @@ -109,7 +109,7 @@ vma_create(struct drm_i915_gem_object *obj, obj->base.size >> PAGE_SHIFT)); vma->size = view->partial.size; vma->size <<= PAGE_SHIFT; - GEM_BUG_ON(vma->size >= obj->base.size); + GEM_BUG_ON(vma->size > obj->base.size); } else if (view->type == I915_GGTT_VIEW_ROTATED) { vma->size = intel_rotation_info_size(&view->rotated); vma->size <<= PAGE_SHIFT; -- cgit v1.2.3