A collection of cleanups
Subsystems (2)
Also includes Nick Hoath's patch to swap context/engine teardown.
In-Reply-To:
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH 1/4 drm/i915: handle teardown of HWSP when context is freed
Existing code did a kunmap() on the wrong page, and didn't account for the fact that the HWSP and the default context are different offsets within the same GEM object.
This patch improves symmetry by creating lrc_teardown_hardware_status_page() to complement lrc_setup_hardware_status_page(), making sure that they both take the same argument (pointer to engine). They encapsulate all the details of the relationship between the default context and the status page, so the rest of the LRC code doesn't have to know about it.
--- drivers/gpu/drm/i915/intel_lrc.c | 57 ++++++++++++++++++++++++++++------------ 1 file changed, 40 insertions(+), 17 deletions(-)
drivers/gpu/drm/i915/intel_lrc.c +40 -17
@@ -226,9 +226,8 @@ enum {#define CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT 0x17static int intel_lr_context_pin (struct drm_i915_gem_request *rq); -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj); - +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring); +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring);/** * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists
@@ -1536,8 +1535,7 @@ static int gen8_init_common_ring(struct intel_engine_cs *ring)struct drm_i915_private *dev_priv = dev->dev_private;u8 next_context_status_buffer_hw; -lrc_setup_hardware_status_page (ring, - dev_priv->kernel_context->engine[ring->id].state); +lrc_setup_hardware_status_page (ring);I915_WRITE_IMR (ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));I915_WRITE (RING_HWSTAM (ring->mmio_base),0xffffffff );
@@ -1992,10 +1990,9 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *ring)i915_cmd_parser_fini_ring (ring);i915_gem_batch_pool_fini (&ring->batch_pool); -if (ring->status_page.obj) { -kunmap (sg_page (ring->status_page.obj->pages->sgl)); - ring->status_page.obj =NULL ; - } +/* Status page should have gone already */ +WARN_ON (ring->status_page.page_addr); +WARN_ON (ring->status_page.obj); ring->disable_lite_restore_wa = false; ring->ctx_desc_template =0 ;
@@ -2434,6 +2431,11 @@ void intel_lr_context_free(struct intel_context *ctx)continue ;if (ctx == ctx->i915->kernel_context) { +/* +* The HWSP is part of the default context +* object in LRC mode. +*/ +lrc_teardown_hardware_status_page (ringbuf->ring);intel_unpin_ringbuffer_obj (ringbuf);i915_gem_object_ggtt_unpin (ctx_obj); }
@@ -2482,24 +2484,45 @@ uint32_t intel_lr_context_size(struct intel_engine_cs *ring)return ret; } -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj) +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring) { -struct drm_i915_private *dev_priv = ring->dev->dev_private; +struct drm_i915_private *dev_priv =to_i915 (ring->dev); +struct intel_context *dctx = dev_priv->kernel_context; +struct drm_i915_gem_object *dctx_obj = dctx->engine[ring->id].state; +u64 dctx_addr =i915_gem_obj_ggtt_offset (dctx_obj);struct page *page; -/* The HWSP is part of the default context object in LRC mode. */ - ring->status_page.gfx_addr =i915_gem_obj_ggtt_offset (default_ctx_obj) - +LRC_PPHWSP_PN *PAGE_SIZE ; - page =i915_gem_object_get_page (default_ctx_obj,LRC_PPHWSP_PN ); +/* +* The HWSP is part of the default context object in LRC mode. +* Note that it doesn't contribute to the refcount! +*/ + page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); ring->status_page.page_addr =kmap (page); - ring->status_page.obj = default_ctx_obj; + ring->status_page.gfx_addr = dctx_addr +LRC_PPHWSP_PN *PAGE_SIZE ; + ring->status_page.obj = dctx_obj;I915_WRITE (RING_HWS_PGA (ring->mmio_base), (u32 )ring->status_page.gfx_addr);POSTING_READ (RING_HWS_PGA (ring->mmio_base)); } +/* This should be called before the default context is destroyed */ +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring) +{ +struct drm_i915_gem_object *dctx_obj = ring->status_page.obj; +struct page *page; + +WARN_ON (!dctx_obj); + +if (ring->status_page.page_addr) { + page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); +kunmap (page); + ring->status_page.page_addr =NULL ; + } + + ring->status_page.obj =NULL ; +} +/** * intel_lr_context_deferred_alloc() - create the LRC specific bits of a context * @ctx: LR context to create.
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 1/4] drm/i915: handle teardown of HWSP when context is freed
Quoted 4 lines · Dave Gordon
Existing code did a
kunmap()on the wrong page, and didn't account for the fact that the HWSP and the default context are different offsets within the same GEM object.
Just to note here that usually we try to split bug fixes early in the series with minimal code changes. This helps cherrypickups to fixes/stable. And also lessens the probability that we accidentally revert bugfix when some other problem occurs with the patch.
Quoted 101 lines
This patch improves symmetry by creating
lrc_teardown_hardware_status_page()to complementlrc_setup_hardware_status_page(), making sure that they both take the same argument (pointer to engine). They encapsulate all the details of the relationship between the default context and the status page, so the rest of the LRC code doesn't have to know about it.Signed-off-by Dave Gordon <[email protected]>--- drivers/gpu/drm/i915/intel_lrc.c | 57 ++++++++++++++++++++++++++++------------ 1 file changed, 40 insertions(+), 17 deletions(-)drivers/gpu/drm/i915/intel_lrc.c +27
-17@@ -226,9 +226,8 @@ enum {#define CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT 0x17static int intel_lr_context_pin (struct drm_i915_gem_request *rq); -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj); - +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring); +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring);/** * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists @@ -1536,8 +1535,7 @@ static int gen8_init_common_ring(struct intel_engine_cs *ring)struct drm_i915_private *dev_priv = dev->dev_private;u8 next_context_status_buffer_hw; -lrc_setup_hardware_status_page (ring, - dev_priv->kernel_context->engine[ring->id].state); +lrc_setup_hardware_status_page (ring);I915_WRITE_IMR (ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));I915_WRITE (RING_HWSTAM (ring->mmio_base),0xffffffff );@@ -1992,10 +1990,9 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *ring)i915_cmd_parser_fini_ring (ring);i915_gem_batch_pool_fini (&ring->batch_pool); -if (ring->status_page.obj) { -kunmap (sg_page (ring->status_page.obj->pages->sgl)); - ring->status_page.obj =NULL ; - } +/* Status page should have gone already */ +WARN_ON (ring->status_page.page_addr); +WARN_ON (ring->status_page.obj); ring->disable_lite_restore_wa = false; ring->ctx_desc_template =0 ;@@ -2434,6 +2431,11 @@ void intel_lr_context_free(struct intel_context *ctx)continue ;if (ctx == ctx->i915->kernel_context) { +/* +* The HWSP is part of the default context +* object in LRC mode. +*/ +lrc_teardown_hardware_status_page (ringbuf->ring);intel_unpin_ringbuffer_obj (ringbuf);i915_gem_object_ggtt_unpin (ctx_obj); }@@ -2482,24 +2484,45 @@ uint32_t intel_lr_context_size(struct intel_engine_cs *ring)return ret; } -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj) +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring) { -struct drm_i915_private *dev_priv = ring->dev->dev_private; +struct drm_i915_private *dev_priv =to_i915 (ring->dev); +struct intel_context *dctx = dev_priv->kernel_context; +struct drm_i915_gem_object *dctx_obj = dctx->engine[ring->id].state; +u64 dctx_addr =i915_gem_obj_ggtt_offset (dctx_obj);struct page *page; -/* The HWSP is part of the default context object in LRC mode. */ - ring->status_page.gfx_addr =i915_gem_obj_ggtt_offset (default_ctx_obj) - +LRC_PPHWSP_PN *PAGE_SIZE ; - page =i915_gem_object_get_page (default_ctx_obj,LRC_PPHWSP_PN ); +/* +* The HWSP is part of the default context object in LRC mode. +* Note that it doesn't contribute to the refcount! +*/ + page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); ring->status_page.page_addr =kmap (page); - ring->status_page.obj = default_ctx_obj; + ring->status_page.gfx_addr = dctx_addr +LRC_PPHWSP_PN *PAGE_SIZE ; + ring->status_page.obj = dctx_obj;I915_WRITE (RING_HWS_PGA (ring->mmio_base), (u32 )ring->status_page.gfx_addr);POSTING_READ (RING_HWS_PGA (ring->mmio_base)); } +/* This should be called before the default context is destroyed */ +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring) +{ +struct drm_i915_gem_object *dctx_obj = ring->status_page.obj;
drivers/gpu/drm/i915/intel_lrc.c uint32_t intel_lr_context_size(struct intel_engine_cs *ring) · line 2522
Quoted 10 lines
drivers/gpu/drm/i915/intel_lrc.c +10
-0+struct page *page; + +WARN_ON (!dctx_obj); + +if (ring->status_page.page_addr) { + page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); +kunmap (page); + ring->status_page.page_addr =NULL ; + } +
As the page is always kmapped along with setting of status_page.obj, I fail to see why we need the check here for the page_addr.
How about just:
if (WARN_ON(!dctx_obj))
return;
-Mika
Quoted 13 lines
drivers/gpu/drm/i915/intel_lrc.c +3
-0+ ring->status_page.obj =NULL ; +} +/** * intel_lr_context_deferred_alloc() - create the LRC specific bits of a context * @ctx: LR context to create. Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 1/4] drm/i915: handle teardown of HWSP when context is freed
Quoted 7 lines · Mika Kuoppala
As the page is always kmapped along with setting of status_page.obj, I fail to see why we need the check here for the
page_addr.How about just:
if (
WARN_ON(!dctx_obj))
return;
Just let it die in a GPF. We already have code to detect impossible situations.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 1/4] drm/i915: handle teardown of HWSP when context is freed
In reply to Mika KuoppalaQuoted 11 lines · Mika Kuoppala
Quoted 4 lines · Dave Gordon
Existing code did a
kunmap()on the wrong page, and didn't account for the fact that the HWSP and the default context are different offsets within the same GEM object.Just to note here that usually we try to split bug fixes early in the series with minimal code changes. This helps cherrypickups to fixes/stable. And also lessens the probability that we accidentally revert bugfix when some other problem occurs with the patch.
OK, let's drop this one from the series (they're orthogonal code changes, only related by the fact they're all to do with init/fini), and I'll resubmit it separately.
Please continue this series from 2/4 ...
Quoted 106 lines
Quoted 106 lines
This patch improves symmetry by creating
lrc_teardown_hardware_status_page()to complementlrc_setup_hardware_status_page(), making sure that they both take the same argument (pointer to engine). They encapsulate all the details of the relationship between the default context and the status page, so the rest of the LRC code doesn't have to know about it.Signed-off-by Dave Gordon <[email protected]>--- drivers/gpu/drm/i915/intel_lrc.c | 57 ++++++++++++++++++++++++++++------------ 1 file changed, 40 insertions(+), 17 deletions(-)drivers/gpu/drm/i915/intel_lrc.c +32
-17@@ -226,9 +226,8 @@ enum {#define CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT 0x17static int intel_lr_context_pin (struct drm_i915_gem_request *rq); -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj); - +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring); +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring);/** * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists @@ -1536,8 +1535,7 @@ static int gen8_init_common_ring(struct intel_engine_cs *ring)struct drm_i915_private *dev_priv = dev->dev_private;u8 next_context_status_buffer_hw; -lrc_setup_hardware_status_page (ring, - dev_priv->kernel_context->engine[ring->id].state); +lrc_setup_hardware_status_page (ring);I915_WRITE_IMR (ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));I915_WRITE (RING_HWSTAM (ring->mmio_base),0xffffffff );@@ -1992,10 +1990,9 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *ring)i915_cmd_parser_fini_ring (ring);i915_gem_batch_pool_fini (&ring->batch_pool); -if (ring->status_page.obj) { -kunmap (sg_page (ring->status_page.obj->pages->sgl)); - ring->status_page.obj =NULL ; - } +/* Status page should have gone already */ +WARN_ON (ring->status_page.page_addr); +WARN_ON (ring->status_page.obj); ring->disable_lite_restore_wa = false; ring->ctx_desc_template =0 ;@@ -2434,6 +2431,11 @@ void intel_lr_context_free(struct intel_context *ctx)continue ;if (ctx == ctx->i915->kernel_context) { +/* +* The HWSP is part of the default context +* object in LRC mode. +*/ +lrc_teardown_hardware_status_page (ringbuf->ring);intel_unpin_ringbuffer_obj (ringbuf);i915_gem_object_ggtt_unpin (ctx_obj); }@@ -2482,24 +2484,45 @@ uint32_t intel_lr_context_size(struct intel_engine_cs *ring)return ret; } -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj) +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring) { -struct drm_i915_private *dev_priv = ring->dev->dev_private; +struct drm_i915_private *dev_priv =to_i915 (ring->dev); +struct intel_context *dctx = dev_priv->kernel_context; +struct drm_i915_gem_object *dctx_obj = dctx->engine[ring->id].state; +u64 dctx_addr =i915_gem_obj_ggtt_offset (dctx_obj);struct page *page; -/* The HWSP is part of the default context object in LRC mode. */ - ring->status_page.gfx_addr =i915_gem_obj_ggtt_offset (default_ctx_obj) - +LRC_PPHWSP_PN *PAGE_SIZE ; - page =i915_gem_object_get_page (default_ctx_obj,LRC_PPHWSP_PN ); +/* +* The HWSP is part of the default context object in LRC mode. +* Note that it doesn't contribute to the refcount! +*/ + page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); ring->status_page.page_addr =kmap (page); - ring->status_page.obj = default_ctx_obj; + ring->status_page.gfx_addr = dctx_addr +LRC_PPHWSP_PN *PAGE_SIZE ; + ring->status_page.obj = dctx_obj;I915_WRITE (RING_HWS_PGA (ring->mmio_base), (u32 )ring->status_page.gfx_addr);POSTING_READ (RING_HWS_PGA (ring->mmio_base)); } +/* This should be called before the default context is destroyed */ +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring) +{ +struct drm_i915_gem_object *dctx_obj = ring->status_page.obj; +struct page *page; + +WARN_ON (!dctx_obj); + +if (ring->status_page.page_addr) {
Quoted 8 lines
drivers/gpu/drm/i915/intel_lrc.c uint32_t intel_lr_context_size(struct intel_engine_cs *ring) · line 2522
Quoted 5 lines
drivers/gpu/drm/i915/intel_lrc.c +5
-0+ page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); +kunmap (page); + ring->status_page.page_addr =NULL ; + } +As the page is always kmapped along with setting of status_page.obj, I fail to see why we need the check here for the
page_addr.
The WARN_ON() tells the developer something unexpected happened.
The page_addr test avoids doing something that may cause an OOPS so that the driver can continue (to unload!).
IMHO, teardown code *in particular* should be written to anticipate being called in all sorts of inconsistent states, because a common error handling strategy is just to say,
oops, something bad happened at some stage of setup! let's tear down everything and pass the error back
which is certainly easier than tracking exactly where the error occurred and undoing only the steps known to have succeeded -- that leads to lots of "goto err_1", "goto err_2", etc :(
This code therefore copes with the situation where the object exists, but the mapping hasn't been set up. With the structure of the code in lrc_setup_hardware_status_page() as implemented above, this won't happen. But someone might someday add a new early error exit that left these variables in that state, and the failure in setup would then result in the driver load failing, at which point teardown() would be called ...
So I might as well guard the kunmap() with a test that the page has actually been kmap()ed, to save some other developer from frustration in the future.
.Dave.
Quoted 16 lines
How about just:
if (
WARN_ON(!dctx_obj))return;-Mika
Quoted 9 lines
drivers/gpu/drm/i915/intel_lrc.c +3
-1+ ring->status_page.obj =NULL ; +} +/** * intel_lr_context_deferred_alloc() - create the LRC specific bits of a context * @ctx: LR context to create. -- 1.9.1
Quoted 4 lines
Quoted 4 lines
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 1/4] drm/i915: handle teardown of HWSP when context is freed
Quoted 143 lines · Dave Gordon
Quoted 11 lines · Mika Kuoppala
Quoted 4 lines · Dave Gordon
Existing code did a
kunmap()on the wrong page, and didn't account for the fact that the HWSP and the default context are different offsets within the same GEM object.Just to note here that usually we try to split bug fixes early in the series with minimal code changes. This helps cherrypickups to fixes/stable. And also lessens the probability that we accidentally revert bugfix when some other problem occurs with the patch.
OK, let's drop this one from the series (they're orthogonal code changes, only related by the fact they're all to do with init/fini), and I'll resubmit it separately.
Please continue this series from 2/4 ...
Quoted 106 lines
Quoted 106 lines
This patch improves symmetry by creating
lrc_teardown_hardware_status_page()to complementlrc_setup_hardware_status_page(), making sure that they both take the same argument (pointer to engine). They encapsulate all the details of the relationship between the default context and the status page, so the rest of the LRC code doesn't have to know about it.Signed-off-by Dave Gordon <[email protected]>--- drivers/gpu/drm/i915/intel_lrc.c | 57 ++++++++++++++++++++++++++++------------ 1 file changed, 40 insertions(+), 17 deletions(-)drivers/gpu/drm/i915/intel_lrc.c +32
-17@@ -226,9 +226,8 @@ enum {#define CTX_RCS_INDIRECT_CTX_OFFSET_DEFAULT 0x17static int intel_lr_context_pin (struct drm_i915_gem_request *rq); -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj); - +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring); +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring);/** * intel_sanitize_enable_execlists() - sanitize i915.enable_execlists @@ -1536,8 +1535,7 @@ static int gen8_init_common_ring(struct intel_engine_cs *ring)struct drm_i915_private *dev_priv = dev->dev_private;u8 next_context_status_buffer_hw; -lrc_setup_hardware_status_page (ring, - dev_priv->kernel_context->engine[ring->id].state); +lrc_setup_hardware_status_page (ring);I915_WRITE_IMR (ring, ~(ring->irq_enable_mask | ring->irq_keep_mask));I915_WRITE (RING_HWSTAM (ring->mmio_base),0xffffffff );@@ -1992,10 +1990,9 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *ring)i915_cmd_parser_fini_ring (ring);i915_gem_batch_pool_fini (&ring->batch_pool); -if (ring->status_page.obj) { -kunmap (sg_page (ring->status_page.obj->pages->sgl)); - ring->status_page.obj =NULL ; - } +/* Status page should have gone already */ +WARN_ON (ring->status_page.page_addr); +WARN_ON (ring->status_page.obj); ring->disable_lite_restore_wa = false; ring->ctx_desc_template =0 ;@@ -2434,6 +2431,11 @@ void intel_lr_context_free(struct intel_context *ctx)continue ;if (ctx == ctx->i915->kernel_context) { +/* +* The HWSP is part of the default context +* object in LRC mode. +*/ +lrc_teardown_hardware_status_page (ringbuf->ring);intel_unpin_ringbuffer_obj (ringbuf);i915_gem_object_ggtt_unpin (ctx_obj); }@@ -2482,24 +2484,45 @@ uint32_t intel_lr_context_size(struct intel_engine_cs *ring)return ret; } -static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring, -struct drm_i915_gem_object *default_ctx_obj) +static void lrc_setup_hardware_status_page (struct intel_engine_cs *ring) { -struct drm_i915_private *dev_priv = ring->dev->dev_private; +struct drm_i915_private *dev_priv =to_i915 (ring->dev); +struct intel_context *dctx = dev_priv->kernel_context; +struct drm_i915_gem_object *dctx_obj = dctx->engine[ring->id].state; +u64 dctx_addr =i915_gem_obj_ggtt_offset (dctx_obj);struct page *page; -/* The HWSP is part of the default context object in LRC mode. */ - ring->status_page.gfx_addr =i915_gem_obj_ggtt_offset (default_ctx_obj) - +LRC_PPHWSP_PN *PAGE_SIZE ; - page =i915_gem_object_get_page (default_ctx_obj,LRC_PPHWSP_PN ); +/* +* The HWSP is part of the default context object in LRC mode. +* Note that it doesn't contribute to the refcount! +*/ + page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); ring->status_page.page_addr =kmap (page); - ring->status_page.obj = default_ctx_obj; + ring->status_page.gfx_addr = dctx_addr +LRC_PPHWSP_PN *PAGE_SIZE ; + ring->status_page.obj = dctx_obj;I915_WRITE (RING_HWS_PGA (ring->mmio_base), (u32 )ring->status_page.gfx_addr);POSTING_READ (RING_HWS_PGA (ring->mmio_base)); } +/* This should be called before the default context is destroyed */ +static void lrc_teardown_hardware_status_page (struct intel_engine_cs *ring) +{ +struct drm_i915_gem_object *dctx_obj = ring->status_page.obj; +struct page *page; + +WARN_ON (!dctx_obj); + +if (ring->status_page.page_addr) {Quoted 8 lines
drivers/gpu/drm/i915/intel_lrc.c uint32_t intel_lr_context_size(struct intel_engine_cs *ring) · line 2522
Quoted 5 lines
drivers/gpu/drm/i915/intel_lrc.c +5
-0+ page =i915_gem_object_get_page (dctx_obj,LRC_PPHWSP_PN ); +kunmap (page); + ring->status_page.page_addr =NULL ; + } +As the page is always kmapped along with setting of status_page.obj, I fail to see why we need the check here for the
page_addr.The
WARN_ON()tells the developer something unexpected happened.The
page_addrtest avoids doing something that may cause an OOPS so that the driver can continue (to unload!).IMHO, teardown code *in particular* should be written to anticipate being called in all sorts of inconsistent states, because a common error handling strategy is just to say,
Quoted 6 lines
oops, something bad happened at some stage of setup! let's tear down everything and pass the error back
which is certainly easier than tracking exactly where the error occurred and undoing only the steps known to have succeeded -- that leads to lots of "goto
err_1", "gotoerr_2", etc :(
Russian dolls error code unwinding is pretty much standard kernel practice. And since it's also mostly dead code (well in practice) it's not too much of a validation horror show. This will likely change a lot if we start using EPROBE_DEFER in earnest.
This comment is just for context, i.e. I think this is totally fine. It's a bit fragile, but you'll be fighting against a _lot_ of inertia with this ;-)
Cheers, Daniel
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH 2/4 drm/i915: Fix context/engine cleanup order
In reply to Dave GordonFrom: Nick Hoath <[email protected]>
Swap the order of context & engine cleanup, so that contexts are cleaned up first, and *then* engines. This is a more sensible order anyway, but in particular has become necessary since the 'intel_ring_initialized() must be simple and inline' patch, which now uses ring->dev as an 'initialised' flag, so it can now be NULL after engine teardown. This in turn can cause a problem in the context code, which (used to) check the ring->dev->struct_mutex -- causing a fault if ring->dev was NULL.
Also rename the cleanup function to reflect what it actually does (cleanup engines, not a ringbuffer), and fix an annoying whitespace issue.
v2: Also make the fix in i915_load_modeset_init, not just in
i915_driver_unload (Chris Wilson)
v3: Had extra stuff in it.
v4: Reverted extra stuff (so we're back to v2).
Rebased and updated commentary above (Dave Gordon).
--- drivers/gpu/drm/i915/i915_dma.c | 4 ++-- drivers/gpu/drm/i915/i915_drv.h | 2 +- drivers/gpu/drm/i915/i915_gem.c | 23 ++++++++++++----------- 3 files changed, 15 insertions(+), 14 deletions(-)
drivers/gpu/drm/i915/i915_dma.c +2 -2
@@ -451,8 +451,8 @@ static int i915_load_modeset_init(struct drm_device *dev) cleanup_gem:mutex_lock (&dev->struct_mutex); -i915_gem_cleanup_ringbuffer (dev);i915_gem_context_fini (dev); +i915_gem_cleanup_engines (dev);mutex_unlock (&dev->struct_mutex); cleanup_irq:intel_guc_ucode_fini (dev);
@@ -1196,8 +1196,8 @@ int i915_driver_unload(struct drm_device *dev)intel_guc_ucode_fini (dev);mutex_lock (&dev->struct_mutex); -i915_gem_cleanup_ringbuffer (dev);i915_gem_context_fini (dev); +i915_gem_cleanup_engines (dev);mutex_unlock (&dev->struct_mutex);intel_fbc_cleanup_cfb (dev_priv);i915_gem_cleanup_stolen (dev);
drivers/gpu/drm/i915/i915_drv.h +1 -1
@@ -3012,7 +3012,7 @@ int i915_gem_init_rings(struct drm_device *dev);int __must_check i915_gem_init_hw (struct drm_device *dev);int i915_gem_l3_remap (struct drm_i915_gem_request *req,int slice);void i915_gem_init_swizzling (struct drm_device *dev); -void i915_gem_cleanup_ringbuffer (struct drm_device *dev); +void i915_gem_cleanup_engines (struct drm_device *dev);int __must_check i915_gpu_idle (struct drm_device *dev);int __must_check i915_gem_suspend (struct drm_device *dev); void __i915_add_request(struct drm_i915_gem_request *req,
drivers/gpu/drm/i915/i915_gem.c +12 -11
@@ -4912,7 +4912,7 @@ int i915_gem_init_rings(struct drm_device *dev) req =i915_gem_request_alloc (ring,NULL );if (IS_ERR (req)) { ret =PTR_ERR (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }
@@ -4925,7 +4925,7 @@ int i915_gem_init_rings(struct drm_device *dev)if (ret && ret != -EIO ) {DRM_ERROR ("PPGTT enable ring #%d failed %d\n" , i, ret);i915_gem_request_cancel (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }
@@ -4933,7 +4933,7 @@ int i915_gem_init_rings(struct drm_device *dev)if (ret && ret != -EIO ) {DRM_ERROR ("Context enable ring #%d failed %d\n" , i, ret);i915_gem_request_cancel (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }
@@ -5008,7 +5008,7 @@ int i915_gem_init(struct drm_device *dev) }void -i915_gem_cleanup_ringbuffer (struct drm_device *dev) +i915_gem_cleanup_engines (struct drm_device *dev) {struct drm_i915_private *dev_priv = dev->dev_private;struct intel_engine_cs *ring;
@@ -5017,13 +5017,14 @@ int i915_gem_init(struct drm_device *dev)for_each_ring (ring, dev_priv, i) dev_priv->gt.cleanup_ring (ring); -if (i915.enable_execlists) -/* -* Neither the BIOS, ourselves or any other kernel -* expects the system to be in execlists mode on startup, -* so we need to reset the GPU back to legacy mode. -*/ -intel_gpu_reset (dev); +if (i915.enable_execlists) { +/* +* Neither the BIOS, ourselves or any other kernel +* expects the system to be in execlists mode on startup, +* so we need to reset the GPU back to legacy mode. +*/ +intel_gpu_reset (dev); + } } static void
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 2/4] drm/i915: Fix context/engine cleanup order
Quoted 19 lines · Dave Gordon
From: Nick Hoath <[email protected]>
Swap the order of context & engine cleanup, so that contexts are cleaned up first, and *then* engines. This is a more sensible order anyway, but in particular has become necessary since the '
intel_ring_initialized()must be simple and inline' patch, which now usesring->devas an 'initialised' flag, so it can now be NULL after engine teardown. This in turn can cause a problem in the context code, which (used to) check thering->dev->struct_mutex-- causing a fault ifring->devwas NULL.Also rename the cleanup function to reflect what it actually does (cleanup engines, not a ringbuffer), and fix an annoying whitespace issue.
v2: Also make the fix in
i915_load_modeset_init, not just in
i915_driver_unload(Chris Wilson)
v3: Had extra stuff in it.
v4: Reverted extra stuff (so we're back to v2).
Rebased and updated commentary above (Dave Gordon).
Quoted 3 lines
Signed-off-by Nick Hoath <[email protected]>Signed-off-by Dave Gordon <[email protected]>Reviewed-by Chris Wilson <[email protected]> (v2)
Trusting that v2==v4 and applied it.
-Daniel
Quoted 113 lines
Cc Mika Kuoppala <[email protected]>Cc Daniel Vetter <[email protected]>Cc Chris Wilson <[email protected]>--- drivers/gpu/drm/i915/i915_dma.c | 4 ++-- drivers/gpu/drm/i915/i915_drv.h | 2 +- drivers/gpu/drm/i915/i915_gem.c | 23 ++++++++++++----------- 3 files changed, 15 insertions(+), 14 deletions(-)drivers/gpu/drm/i915/i915_dma.c +2
-2@@ -451,8 +451,8 @@ static int i915_load_modeset_init(struct drm_device *dev) cleanup_gem:mutex_lock (&dev->struct_mutex); -i915_gem_cleanup_ringbuffer (dev);i915_gem_context_fini (dev); +i915_gem_cleanup_engines (dev);mutex_unlock (&dev->struct_mutex); cleanup_irq:intel_guc_ucode_fini (dev);@@ -1196,8 +1196,8 @@ int i915_driver_unload(struct drm_device *dev)intel_guc_ucode_fini (dev);mutex_lock (&dev->struct_mutex); -i915_gem_cleanup_ringbuffer (dev);i915_gem_context_fini (dev); +i915_gem_cleanup_engines (dev);mutex_unlock (&dev->struct_mutex);intel_fbc_cleanup_cfb (dev_priv);i915_gem_cleanup_stolen (dev);drivers/gpu/drm/i915/i915_drv.h +1
-1@@ -3012,7 +3012,7 @@ int i915_gem_init_rings(struct drm_device *dev);int __must_check i915_gem_init_hw (struct drm_device *dev);int i915_gem_l3_remap (struct drm_i915_gem_request *req,int slice);void i915_gem_init_swizzling (struct drm_device *dev); -void i915_gem_cleanup_ringbuffer (struct drm_device *dev); +void i915_gem_cleanup_engines (struct drm_device *dev);int __must_check i915_gpu_idle (struct drm_device *dev);int __must_check i915_gem_suspend (struct drm_device *dev); void __i915_add_request(struct drm_i915_gem_request *req,drivers/gpu/drm/i915/i915_gem.c +12
-11@@ -4912,7 +4912,7 @@ int i915_gem_init_rings(struct drm_device *dev) req =i915_gem_request_alloc (ring,NULL );if (IS_ERR (req)) { ret =PTR_ERR (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }@@ -4925,7 +4925,7 @@ int i915_gem_init_rings(struct drm_device *dev)if (ret && ret != -EIO ) {DRM_ERROR ("PPGTT enable ring #%d failed %d\n" , i, ret);i915_gem_request_cancel (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }@@ -4933,7 +4933,7 @@ int i915_gem_init_rings(struct drm_device *dev)if (ret && ret != -EIO ) {DRM_ERROR ("Context enable ring #%d failed %d\n" , i, ret);i915_gem_request_cancel (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }@@ -5008,7 +5008,7 @@ int i915_gem_init(struct drm_device *dev) }void -i915_gem_cleanup_ringbuffer (struct drm_device *dev) +i915_gem_cleanup_engines (struct drm_device *dev) {struct drm_i915_private *dev_priv = dev->dev_private;struct intel_engine_cs *ring;@@ -5017,13 +5017,14 @@ int i915_gem_init(struct drm_device *dev)for_each_ring (ring, dev_priv, i) dev_priv->gt.cleanup_ring (ring); -if (i915.enable_execlists) -/* -* Neither the BIOS, ourselves or any other kernel -* expects the system to be in execlists mode on startup, -* so we need to reset the GPU back to legacy mode. -*/ -intel_gpu_reset (dev); +if (i915.enable_execlists) { +/* +* Neither the BIOS, ourselves or any other kernel +* expects the system to be in execlists mode on startup, +* so we need to reset the GPU back to legacy mode. +*/ +intel_gpu_reset (dev); + } } static voidSignature · 1 line
1.9.1
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 2/4] drm/i915: Fix context/engine cleanup order
In reply to Dave GordonQuoted 48 lines · Dave Gordon
From: Nick Hoath <[email protected]>
Swap the order of context & engine cleanup, so that contexts are cleaned up first, and *then* engines. This is a more sensible order anyway, but in particular has become necessary since the '
intel_ring_initialized()must be simple and inline' patch, which now usesring->devas an 'initialised' flag, so it can now be NULL after engine teardown. This in turn can cause a problem in the context code, which (used to) check thering->dev->struct_mutex-- causing a fault ifring->devwas NULL.Also rename the cleanup function to reflect what it actually does (cleanup engines, not a ringbuffer), and fix an annoying whitespace issue.
v2: Also make the fix in
i915_load_modeset_init, not just in
i915_driver_unload(Chris Wilson)
v3: Had extra stuff in it.
v4: Reverted extra stuff (so we're back to v2).
Rebased and updated commentary above (Dave Gordon).Signed-off-by Nick Hoath <[email protected]>Signed-off-by Dave Gordon <[email protected]>Reviewed-by Chris Wilson <[email protected]> (v2)Cc Mika Kuoppala <[email protected]>Cc Daniel Vetter <[email protected]>Cc Chris Wilson <[email protected]>--- drivers/gpu/drm/i915/i915_dma.c | 4 ++-- drivers/gpu/drm/i915/i915_drv.h | 2 +- drivers/gpu/drm/i915/i915_gem.c | 23 ++++++++++++----------- 3 files changed, 15 insertions(+), 14 deletions(-)drivers/gpu/drm/i915/i915_dma.c +1
-1@@ -451,8 +451,8 @@ static int i915_load_modeset_init(struct drm_device *dev) cleanup_gem:mutex_lock (&dev->struct_mutex); -i915_gem_cleanup_ringbuffer (dev);i915_gem_context_fini (dev); +i915_gem_cleanup_engines (dev);mutex_unlock (&dev->struct_mutex); cleanup_irq:intel_guc_ucode_fini (dev);
drivers/gpu/drm/i915/i915_dma.c int i915_driver_unload(struct drm_device *dev) · line 1200
Quoted 6 lines
drivers/gpu/drm/i915/i915_dma.c +1
-1@@ -1196,8 +1196,8 @@ int i915_driver_unload(struct drm_device *dev)intel_guc_ucode_fini (dev);mutex_lock (&dev->struct_mutex); -i915_gem_cleanup_ringbuffer (dev);i915_gem_context_fini (dev); +i915_gem_cleanup_engines (dev);
context_fini assumes it is the one who is dropping the last reference to the default/kernel context and NULLs the pointer.
Going forward this means there is this hidden dependency that engines must not hold the reference to the kernel context, at least not rely on dev_priv->kernel_context, or we would either leak the reference or null ptr deref.
I don't know, somehow feels a bit fragile. But a lot of things seem fragile around here.
If the main motivation is to ensure access to ring->dev, then ctx->i915 could be used from the context cleanup code. For the ordering itself I am not sure what is more logical. Idle GPU -> shutdown engines -> cleanup contexts sounds OK to me since the engines are the ones which can hold references to contexts. So logically cleaning up context before engines feels strange. But I am sure an opposite argument could also be made?
Regards,
Tvrtko
Quoted 69 lines
drivers/gpu/drm/i915/i915_dma.c +0
-0mutex_unlock (&dev->struct_mutex);intel_fbc_cleanup_cfb (dev_priv);i915_gem_cleanup_stolen (dev);drivers/gpu/drm/i915/i915_drv.h +1
-1@@ -3012,7 +3012,7 @@ int i915_gem_init_rings(struct drm_device *dev);int __must_check i915_gem_init_hw (struct drm_device *dev);int i915_gem_l3_remap (struct drm_i915_gem_request *req,int slice);void i915_gem_init_swizzling (struct drm_device *dev); -void i915_gem_cleanup_ringbuffer (struct drm_device *dev); +void i915_gem_cleanup_engines (struct drm_device *dev);int __must_check i915_gpu_idle (struct drm_device *dev);int __must_check i915_gem_suspend (struct drm_device *dev); void __i915_add_request(struct drm_i915_gem_request *req,drivers/gpu/drm/i915/i915_gem.c +5
-11@@ -4912,7 +4912,7 @@ int i915_gem_init_rings(struct drm_device *dev) req =i915_gem_request_alloc (ring,NULL );if (IS_ERR (req)) { ret =PTR_ERR (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }@@ -4925,7 +4925,7 @@ int i915_gem_init_rings(struct drm_device *dev)if (ret && ret != -EIO ) {DRM_ERROR ("PPGTT enable ring #%d failed %d\n" , i, ret);i915_gem_request_cancel (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }@@ -4933,7 +4933,7 @@ int i915_gem_init_rings(struct drm_device *dev)if (ret && ret != -EIO ) {DRM_ERROR ("Context enable ring #%d failed %d\n" , i, ret);i915_gem_request_cancel (req); -i915_gem_cleanup_ringbuffer (dev); +i915_gem_cleanup_engines (dev); goto out; }@@ -5008,7 +5008,7 @@ int i915_gem_init(struct drm_device *dev) }void -i915_gem_cleanup_ringbuffer (struct drm_device *dev) +i915_gem_cleanup_engines (struct drm_device *dev) {struct drm_i915_private *dev_priv = dev->dev_private;struct intel_engine_cs *ring;@@ -5017,13 +5017,14 @@ int i915_gem_init(struct drm_device *dev)for_each_ring (ring, dev_priv, i) dev_priv->gt.cleanup_ring (ring); -if (i915.enable_execlists) -/* -* Neither the BIOS, ourselves or any other kernel -* expects the system to be in execlists mode on startup, -* so we need to reset the GPU back to legacy mode. -*/ -intel_gpu_reset (dev); +if (i915.enable_execlists) {
drivers/gpu/drm/i915/i915_gem.c int i915_gem_init(struct drm_device *dev) · line 5030
Quoted 10 lines
drivers/gpu/drm/i915/i915_gem.c +7
-0+/* +* Neither the BIOS, ourselves or any other kernel +* expects the system to be in execlists mode on startup, +* so we need to reset the GPU back to legacy mode. +*/ +intel_gpu_reset (dev); + } } static void
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH 3/4 drm/i915: tidy up initialisation failure paths (legacy)
In reply to Dave Gordon1. Fix intel_cleanup_ring_buffer() to handle the error cleanup case where the ringbuffer has been allocated but map-and-pin failed. Unpin it iff it's previously been mapped-and-pinned.
2. Fix the error path in intel_init_ring_buffer(), which already called intel_destroy_ringbuffer_obj(), but failed to free the actual ringbuffer structure. Calling intel_ringbuffer_free() instead does both in one go.
3. With the above change, intel_destroy_ringbuffer_obj() is only called in one place (intel_ringbuffer_free()), so flatten it into that function.
4. move low-level register accesses from intel_cleanup_ring_buffer() (which calls intel_stop_ring_buffer(ring) which calls stop_ring()) down into stop_ring() itself), which is already doing low-level register accesses. Then, intel_cleanup_ring_buffer() no longer needs 'dev_priv'.
--- drivers/gpu/drm/i915/intel_ringbuffer.c | 47 +++++++++++++++------------------ 1 file changed, 22 insertions(+), 25 deletions(-)
drivers/gpu/drm/i915/intel_ringbuffer.c +22 -25
@@ -551,6 +551,8 @@ static bool stop_ring(struct intel_engine_cs *ring)I915_WRITE_MODE (ring,_MASKED_BIT_DISABLE (STOP_RING )); } +WARN_ON (!IS_GEN2 (ring->dev) && (I915_READ_MODE (ring) &MODE_IDLE ) ==0 ); +return (I915_READ_HEAD (ring) &HEAD_ADDR ) ==0 ; }
@@ -2071,12 +2073,6 @@ int intel_pin_and_map_ringbuffer_obj(struct drm_device *dev,return 0 ; } -static void intel_destroy_ringbuffer_obj (struct intel_ringbuffer *ringbuf) -{ -drm_gem_object_unreference (&ringbuf->obj->base); - ringbuf->obj =NULL ; -} -static int intel_alloc_ringbuffer_obj (struct drm_device *dev,struct intel_ringbuffer *ringbuf) {
@@ -2139,11 +2135,14 @@ struct intel_ringbuffer * }void -intel_ringbuffer_free (struct intel_ringbuffer *ring) +intel_ringbuffer_free (struct intel_ringbuffer *ringbuf) { -intel_destroy_ringbuffer_obj (ring); -list_del (&ring->link); -kfree (ring); +if (ringbuf->obj) { +drm_gem_object_unreference (&ringbuf->obj->base); + ringbuf->obj =NULL ; + } +list_del (&ringbuf->link); +kfree (ringbuf); } static int intel_init_ring_buffer(struct drm_device *dev,
@@ -2171,6 +2170,13 @@ static int intel_init_ring_buffer(struct drm_device *dev, } ring->buffer = ringbuf; + ret =intel_pin_and_map_ringbuffer_obj (dev, ringbuf); +if (ret) { +DRM_ERROR ("Failed to pin and map ringbuffer %s: %d\n" , + ring->name, ret); + goto error; + } +if (I915_NEED_GFX_HWS (dev)) { ret =init_status_page (ring);if (ret)
@@ -2182,14 +2188,6 @@ static int intel_init_ring_buffer(struct drm_device *dev, goto error; } - ret =intel_pin_and_map_ringbuffer_obj (dev, ringbuf); -if (ret) { -DRM_ERROR ("Failed to pin and map ringbuffer %s: %d\n" , - ring->name, ret); -intel_destroy_ringbuffer_obj (ringbuf); - goto error; - } - ret =i915_cmd_parser_init_ring (ring);if (ret) goto error;
@@ -2203,19 +2201,18 @@ static int intel_init_ring_buffer(struct drm_device *dev,void intel_cleanup_ring_buffer (struct intel_engine_cs *ring) { -struct drm_i915_private *dev_priv; +struct intel_ringbuffer *ringbuf;if (!intel_ring_initialized (ring))return ; - dev_priv =to_i915 (ring->dev); - -if (ring->buffer) { + ringbuf = ring->buffer; +if (ringbuf) {intel_stop_ring_buffer (ring); -WARN_ON (!IS_GEN2 (ring->dev) && (I915_READ_MODE (ring) &MODE_IDLE ) ==0 ); -intel_unpin_ringbuffer_obj (ring->buffer); -intel_ringbuffer_free (ring->buffer); +if (ringbuf->virtual_start) +intel_unpin_ringbuffer_obj (ringbuf); +intel_ringbuffer_free (ringbuf); ring->buffer =NULL ; }
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
PATCH 4/4 drm/i915: tidy up initialisation failure paths (GEM & LRC)
In reply to Dave Gordon1. add call to i915_gem_context_fini() to deallocate the default context(s) if the call to init_rings() fails, so that we don't leak the context in that situation.
2. remove useless code in intel_logical_ring_cleanup(), presumably copypasted from legacy ringbuffer version at creation.
--- drivers/gpu/drm/i915/i915_gem.c | 5 ++++- drivers/gpu/drm/i915/intel_lrc.c | 10 ++-------- 2 files changed, 6 insertions(+), 9 deletions(-)
drivers/gpu/drm/i915/i915_gem.c +4 -1
@@ -4986,8 +4986,11 @@ int i915_gem_init(struct drm_device *dev) goto out_unlock; ret = dev_priv->gt.init_rings (dev); -if (ret) +if (ret) { +i915_gem_context_fini (dev); +/* XXX: anything else to be undone here? */ goto out_unlock; + } ret =i915_gem_init_hw (dev);if (ret == -EIO ) {
drivers/gpu/drm/i915/intel_lrc.c +2 -8
@@ -1972,17 +1972,11 @@ static int gen8_init_rcs_context(struct drm_i915_gem_request *req)*/ void intel_logical_ring_cleanup (struct intel_engine_cs *ring) { -struct drm_i915_private *dev_priv; -if (!intel_ring_initialized (ring))return ; - dev_priv = ring->dev->dev_private; - -if (ring->buffer) { -intel_logical_ring_stop (ring); -WARN_ON ((I915_READ_MODE (ring) &MODE_IDLE ) ==0 ); - } +/* should not be set in LRC mode */ +WARN_ON (ring->buffer);if (ring->cleanup) ring->cleanup (ring);
Signature · 6 lines
1.9.1 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 4 lines · Dave Gordon
Also includes Nick Hoath's patch to swap context/engine teardown.
In-Reply-To:
Hmm, editor/mailer seems to have dropped most of this summary :(
Probably something to do with that IRT-line above ... what it was supposed to say was approximately this:
Subject: [PATCH 0/4] A collection of cleanups
Various things can go wrong during initialisation and teardown, but they usually don't, so the error-handling paths go largely untested. This collection of patches fixes some things I recently noticed. Some might lead to a kernel OOPS, but mostly they're leaks and other inconsistencies.
Also includes Nick Hoath's patch to swap context/engine teardown, because that helps make setup & teardown more consistent too.
.Dave.
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✗ Fi.CI.BAT: warning for series starting with [1/4] drm/i915: handle teardown of HWSP when context is freed
In reply to Dave Gordon== Summary ==
== Summary ==
Built on 8fe9e785ae04fa7c37f7935cff12d62e38054b60 drm-intel-nightly: 2016y-01m-21d-11h-02m-42s UTC integration manifest
Test kms_pipe_crc_basic:
Subgroup read-crc-pipe-a-frame-sequence:
pass -> DMESG-WARN (ilk-hp8440p)bdw-nuci7 total:140 pass:131 dwarn:0 dfail:0 fail:0 skip:9 bdw-ultra total:143 pass:137 dwarn:0 dfail:0 fail:0 skip:6 bsw-nuc-2 total:143 pass:119 dwarn:0 dfail:0 fail:0 skip:24 byt-nuc total:143 pass:128 dwarn:0 dfail:0 fail:0 skip:15 hsw-brixbox total:143 pass:136 dwarn:0 dfail:0 fail:0 skip:7 hsw-gt2 total:143 pass:139 dwarn:0 dfail:0 fail:0 skip:4 ilk-hp8440p total:143 pass:103 dwarn:2 dfail:0 fail:0 skip:38 ivb-t430s total:143 pass:137 dwarn:0 dfail:0 fail:0 skip:6 skl-i5k-2 total:143 pass:134 dwarn:1 dfail:0 fail:0 skip:8 skl-i7k-2 total:143 pass:134 dwarn:1 dfail:0 fail:0 skip:8 snb-dellxps total:143 pass:129 dwarn:0 dfail:0 fail:0 skip:14 snb-x220t total:143 pass:129 dwarn:0 dfail:0 fail:1 skip:13
Results at /archive/results/CI_IGT_test/Patchwork_1241/
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: ✗ Fi.CI.BAT: warning for series starting with [1/4] drm/i915: handle teardown of HWSP when context is freed
Quoted 7 lines · Patchwork
== Summary ==
Built on
8fe9e785ae04fa7c37f7935cff12d62e38054b60drm-intel-nightly: 2016y-01m-21d-11h-02m-42s UTC integration manifestTest
kms_pipe_crc_basic:Subgroup read-crc-pipe-a-frame-sequence: pass -> DMESG-WARN (ilk-hp8440p)
https://bugs.freedesktop.org/show_bug.cgi?id=93787
"[BAT ILK] sporadic fifo underruns" (again)
.Dave.
Quoted 13 lines
bdw-nuci7 total:140 pass:131 dwarn:0 dfail:0 fail:0 skip:9 bdw-ultra total:143 pass:137 dwarn:0 dfail:0 fail:0 skip:6 bsw-nuc-2 total:143 pass:119 dwarn:0 dfail:0 fail:0 skip:24 byt-nuc total:143 pass:128 dwarn:0 dfail:0 fail:0 skip:15 hsw-brixbox total:143 pass:136 dwarn:0 dfail:0 fail:0 skip:7 hsw-gt2 total:143 pass:139 dwarn:0 dfail:0 fail:0 skip:4 ilk-hp8440p total:143 pass:103 dwarn:2 dfail:0 fail:0 skip:38 ivb-t430s total:143 pass:137 dwarn:0 dfail:0 fail:0 skip:6 skl-i5k-2 total:143 pass:134 dwarn:1 dfail:0 fail:0 skip:8 skl-i7k-2 total:143 pass:134 dwarn:1 dfail:0 fail:0 skip:8 snb-dellxps total:143 pass:129 dwarn:0 dfail:0 fail:0 skip:14 snb-x220t total:143 pass:129 dwarn:0 dfail:0 fail:1 skip:13
Quoted 1 line
Results at /archive/results/CI_IGT_test/Patchwork_1241/
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx