| Message ID | 20241119135030.3352939-1-karunika.choo@arm.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | [v2] drm/panthor: Simplify FW fast reset path | expand |
On Tue, 19 Nov 2024 13:50:29 +0000 Karunika Choo <karunika.choo@arm.com> wrote: > Stop checking the FW halt_status as MCU_STATUS should be sufficient. > This should make the check for successful FW halt and subsequently > setting fast_reset to true more robust. > > We should also clear GLB_REQ.GLB_HALT bit only on post-reset prior > to starting the FW and only if we're doing a fast reset, because > the slow reset will re-initialize all FW sections, including the > global interface. > > Signed-off-by: Karunika Choo <karunika.choo@arm.com> Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com> and I think you forgot to add Liviu's R-b > --- > v2: > - clarify comments and commit message with regards to when to clear the > GLB_HALT flag. > > drivers/gpu/drm/panthor/panthor_fw.c | 15 ++++++++------- > 1 file changed, 8 insertions(+), 7 deletions(-) > > diff --git a/drivers/gpu/drm/panthor/panthor_fw.c b/drivers/gpu/drm/panthor/panthor_fw.c > index ecca5565ce41..4bc52b1b1a28 100644 > --- a/drivers/gpu/drm/panthor/panthor_fw.c > +++ b/drivers/gpu/drm/panthor/panthor_fw.c > @@ -1098,17 +1098,11 @@ void panthor_fw_pre_reset(struct panthor_device *ptdev, bool on_hang) > panthor_fw_update_reqs(glb_iface, req, GLB_HALT, GLB_HALT); > gpu_write(ptdev, CSF_DOORBELL(CSF_GLB_DOORBELL_ID), 1); > if (!readl_poll_timeout(ptdev->iomem + MCU_STATUS, status, > - status == MCU_STATUS_HALT, 10, 100000) && > - glb_iface->output->halt_status == PANTHOR_FW_HALT_OK) { > + status == MCU_STATUS_HALT, 10, 100000)) { > ptdev->fw->fast_reset = true; > } else { > drm_warn(&ptdev->base, "Failed to cleanly suspend MCU"); > } > - > - /* The FW detects 0 -> 1 transitions. Make sure we reset > - * the HALT bit before the FW is rebooted. > - */ > - panthor_fw_update_reqs(glb_iface, req, 0, GLB_HALT); > } > > panthor_job_irq_suspend(&ptdev->fw->irq); > @@ -1134,6 +1128,13 @@ int panthor_fw_post_reset(struct panthor_device *ptdev) > * the FW sections. If it fails, go for a full reset. > */ > if (ptdev->fw->fast_reset) { > + /* The FW detects 0 -> 1 transitions. Make sure we reset > + * the HALT bit before the FW is rebooted. > + * This is not needed on a slow reset because FW sections are > + * re-initialized. > + */ > + panthor_fw_update_reqs(glb_iface, req, 0, GLB_HALT); > + > ret = panthor_fw_start(ptdev); > if (!ret) > goto out;
On Tue, Nov 19, 2024 at 03:25:25PM +0100, Boris Brezillon wrote: > On Tue, 19 Nov 2024 13:50:29 +0000 > Karunika Choo <karunika.choo@arm.com> wrote: > > > Stop checking the FW halt_status as MCU_STATUS should be sufficient. > > This should make the check for successful FW halt and subsequently > > setting fast_reset to true more robust. > > > > We should also clear GLB_REQ.GLB_HALT bit only on post-reset prior > > to starting the FW and only if we're doing a fast reset, because > > the slow reset will re-initialize all FW sections, including the > > global interface. > > > > Signed-off-by: Karunika Choo <karunika.choo@arm.com> > > Reviewed-by: Boris Brezillon <boris.brezillon@collabora.com> > > and I think you forgot to add Liviu's R-b I've fixed the commit and pushed it without compile testing it because I did it for the patch before it. Obviously it doesn't compile because glb_iface is not defined :( Adding a new patch, can you guys R-b this? diff --git a/drivers/gpu/drm/panthor/panthor_fw.c b/drivers/gpu/drm/panthor/panthor_fw.c index 4bc52b1b1a286..c807b6ce71bd4 100644 --- a/drivers/gpu/drm/panthor/panthor_fw.c +++ b/drivers/gpu/drm/panthor/panthor_fw.c @@ -1133,6 +1133,7 @@ int panthor_fw_post_reset(struct panthor_device *ptdev) * This is not needed on a slow reset because FW sections are * re-initialized. */ + struct panthor_fw_global_iface *glb_iface = panthor_fw_get_glb_iface(ptdev); panthor_fw_update_reqs(glb_iface, req, 0, GLB_HALT); ret = panthor_fw_start(ptdev); Best regards, Liviu > > > --- > > v2: > > - clarify comments and commit message with regards to when to clear the > > GLB_HALT flag. > > > > drivers/gpu/drm/panthor/panthor_fw.c | 15 ++++++++------- > > 1 file changed, 8 insertions(+), 7 deletions(-) > > > > diff --git a/drivers/gpu/drm/panthor/panthor_fw.c b/drivers/gpu/drm/panthor/panthor_fw.c > > index ecca5565ce41..4bc52b1b1a28 100644 > > --- a/drivers/gpu/drm/panthor/panthor_fw.c > > +++ b/drivers/gpu/drm/panthor/panthor_fw.c > > @@ -1098,17 +1098,11 @@ void panthor_fw_pre_reset(struct panthor_device *ptdev, bool on_hang) > > panthor_fw_update_reqs(glb_iface, req, GLB_HALT, GLB_HALT); > > gpu_write(ptdev, CSF_DOORBELL(CSF_GLB_DOORBELL_ID), 1); > > if (!readl_poll_timeout(ptdev->iomem + MCU_STATUS, status, > > - status == MCU_STATUS_HALT, 10, 100000) && > > - glb_iface->output->halt_status == PANTHOR_FW_HALT_OK) { > > + status == MCU_STATUS_HALT, 10, 100000)) { > > ptdev->fw->fast_reset = true; > > } else { > > drm_warn(&ptdev->base, "Failed to cleanly suspend MCU"); > > } > > - > > - /* The FW detects 0 -> 1 transitions. Make sure we reset > > - * the HALT bit before the FW is rebooted. > > - */ > > - panthor_fw_update_reqs(glb_iface, req, 0, GLB_HALT); > > } > > > > panthor_job_irq_suspend(&ptdev->fw->irq); > > @@ -1134,6 +1128,13 @@ int panthor_fw_post_reset(struct panthor_device *ptdev) > > * the FW sections. If it fails, go for a full reset. > > */ > > if (ptdev->fw->fast_reset) { > > + /* The FW detects 0 -> 1 transitions. Make sure we reset > > + * the HALT bit before the FW is rebooted. > > + * This is not needed on a slow reset because FW sections are > > + * re-initialized. > > + */ > > + panthor_fw_update_reqs(glb_iface, req, 0, GLB_HALT); > > + > > ret = panthor_fw_start(ptdev); > > if (!ret) > > goto out; >
diff --git a/drivers/gpu/drm/panthor/panthor_fw.c b/drivers/gpu/drm/panthor/panthor_fw.c index ecca5565ce41..4bc52b1b1a28 100644 --- a/drivers/gpu/drm/panthor/panthor_fw.c +++ b/drivers/gpu/drm/panthor/panthor_fw.c @@ -1098,17 +1098,11 @@ void panthor_fw_pre_reset(struct panthor_device *ptdev, bool on_hang) panthor_fw_update_reqs(glb_iface, req, GLB_HALT, GLB_HALT); gpu_write(ptdev, CSF_DOORBELL(CSF_GLB_DOORBELL_ID), 1); if (!readl_poll_timeout(ptdev->iomem + MCU_STATUS, status, - status == MCU_STATUS_HALT, 10, 100000) && - glb_iface->output->halt_status == PANTHOR_FW_HALT_OK) { + status == MCU_STATUS_HALT, 10, 100000)) { ptdev->fw->fast_reset = true; } else { drm_warn(&ptdev->base, "Failed to cleanly suspend MCU"); } - - /* The FW detects 0 -> 1 transitions. Make sure we reset - * the HALT bit before the FW is rebooted. - */ - panthor_fw_update_reqs(glb_iface, req, 0, GLB_HALT); } panthor_job_irq_suspend(&ptdev->fw->irq); @@ -1134,6 +1128,13 @@ int panthor_fw_post_reset(struct panthor_device *ptdev) * the FW sections. If it fails, go for a full reset. */ if (ptdev->fw->fast_reset) { + /* The FW detects 0 -> 1 transitions. Make sure we reset + * the HALT bit before the FW is rebooted. + * This is not needed on a slow reset because FW sections are + * re-initialized. + */ + panthor_fw_update_reqs(glb_iface, req, 0, GLB_HALT); + ret = panthor_fw_start(ptdev); if (!ret) goto out;
Stop checking the FW halt_status as MCU_STATUS should be sufficient. This should make the check for successful FW halt and subsequently setting fast_reset to true more robust. We should also clear GLB_REQ.GLB_HALT bit only on post-reset prior to starting the FW and only if we're doing a fast reset, because the slow reset will re-initialize all FW sections, including the global interface. Signed-off-by: Karunika Choo <karunika.choo@arm.com> --- v2: - clarify comments and commit message with regards to when to clear the GLB_HALT flag. drivers/gpu/drm/panthor/panthor_fw.c | 15 ++++++++------- 1 file changed, 8 insertions(+), 7 deletions(-)