| Message ID | 20200406221253.1307209-4-lyude@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | drm/dp_mst: drm_dp_check_act_status() fixes | expand |
On Mon, Apr 6, 2020 at 6:13 PM Lyude Paul <lyude@redhat.com> wrote: > > Currently we only poll for an ACT up to 30 times, with a busy-wait delay > of 100µs between each attempt - giving us a timeout of 2900µs. While > this might seem sensible, it would appear that in certain scenarios it > can take dramatically longer then that for us to receive an ACT. On one > of the EVGA MST hubs that I have available, I observed said hub > sometimes taking longer then a second before signalling the ACT. These > delays mostly seem to occur when previous sideband messages we've sent > are NAKd by the hub, however it wouldn't be particularly surprising if > it's possible to reproduce times like this simply by introducing branch > devices with large LCTs since payload allocations have to take effect on > every downstream device up to the payload's target. > > So, instead of just retrying 30 times we poll for the ACT for up to 3ms, > and additionally use usleep_range() to avoid a very long and rude > busy-wait. Note that the previous retry count of 30 appears to have been > arbitrarily chosen, as I can't find any mention of a recommended timeout > or retry count for ACTs in the DisplayPort 2.0 specification. This also > goes for the range we were previously using for udelay(), although I > suspect that was just copied from the recommended delay for link > training on SST devices. > > Changes since v1: > * Use readx_poll_timeout() instead of open-coding timeout loop - Sean > Paul > > Signed-off-by: Lyude Paul <lyude@redhat.com> > Fixes: ad7f8a1f9ced ("drm/helper: add Displayport multi-stream helper (v0.6)") > Cc: Sean Paul <sean@poorly.run> > Cc: <stable@vger.kernel.org> # v3.17+ > --- > drivers/gpu/drm/drm_dp_mst_topology.c | 57 ++++++++++++++++----------- > 1 file changed, 34 insertions(+), 23 deletions(-) > > diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c > index c83adbdfc1cd..ce61964baa7c 100644 > --- a/drivers/gpu/drm/drm_dp_mst_topology.c > +++ b/drivers/gpu/drm/drm_dp_mst_topology.c > @@ -27,6 +27,7 @@ > #include <linux/kernel.h> > #include <linux/sched.h> > #include <linux/seq_file.h> > +#include <linux/iopoll.h> > > #if IS_ENABLED(CONFIG_DRM_DEBUG_DP_MST_TOPOLOGY_REFS) > #include <linux/stacktrace.h> > @@ -4460,43 +4461,53 @@ static int drm_dp_dpcd_write_payload(struct drm_dp_mst_topology_mgr *mgr, > return ret; > } > > +static int do_get_act_status(struct drm_dp_aux *aux) > +{ > + int ret; > + u8 status; > + > + ret = drm_dp_dpcd_readb(aux, DP_PAYLOAD_TABLE_UPDATE_STATUS, &status); > + if (ret < 0) > + return ret; > + > + return status; > +} > > /** > * drm_dp_check_act_status() - Polls for ACT handled status. > * @mgr: manager to use > * > * Tries waiting for the MST hub to finish updating it's payload table by > - * polling for the ACT handled bit. > + * polling for the ACT handled bit for up to 3 seconds (yes-some hubs really > + * take that long). > * > * Returns: > * 0 if the ACT was handled in time, negative error code on failure. > */ > int drm_dp_check_act_status(struct drm_dp_mst_topology_mgr *mgr) > { > - int count = 0, ret; > - u8 status; > - > - do { > - ret = drm_dp_dpcd_readb(mgr->aux, > - DP_PAYLOAD_TABLE_UPDATE_STATUS, > - &status); > - if (ret < 0) { > - DRM_DEBUG_KMS("failed to read payload table status %d\n", > - ret); > - return ret; > - } > - > - if (status & DP_PAYLOAD_ACT_HANDLED) > - break; > - count++; > - udelay(100); > - } while (count < 30); > - > - if (!(status & DP_PAYLOAD_ACT_HANDLED)) { > - DRM_DEBUG_KMS("failed to get ACT bit %d after %d retries\n", > - status, count); > + /* > + * There doesn't seem to be any recommended retry count or timeout in > + * the MST specification. Since some hubs have been observed to take > + * over 1 second to update their payload allocations under certain > + * conditions, we use a rather large timeout value. > + */ > + const int timeout_ms = 3000; > + int ret, status; > + > + ret = readx_poll_timeout(do_get_act_status, mgr->aux, status, > + status & DP_PAYLOAD_ACT_HANDLED || status < 0, > + 100, timeout_ms * USEC_PER_MSEC); In v1 the usleep range was 100 -> 1000, in v2 it's going to be 51 -> 100. Perhaps bump this up to 200? > + if (ret < 0 && status >= 0) { > + DRM_DEBUG_KMS("Failed to get ACT bit %d after %dms\n", > + status, timeout_ms); I still think status should be base 16 when printed With those nits addressed, Reviewed-by: Sean Paul <sean@poorly.run> > return -EINVAL; > + } else if (status < 0) { > + DRM_DEBUG_KMS("Failed to read payload table status: %d\n", > + status); > + return status; > } > + > return 0; > } > EXPORT_SYMBOL(drm_dp_check_act_status); > -- > 2.25.1 >
diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c index c83adbdfc1cd..ce61964baa7c 100644 --- a/drivers/gpu/drm/drm_dp_mst_topology.c +++ b/drivers/gpu/drm/drm_dp_mst_topology.c @@ -27,6 +27,7 @@ #include <linux/kernel.h> #include <linux/sched.h> #include <linux/seq_file.h> +#include <linux/iopoll.h> #if IS_ENABLED(CONFIG_DRM_DEBUG_DP_MST_TOPOLOGY_REFS) #include <linux/stacktrace.h> @@ -4460,43 +4461,53 @@ static int drm_dp_dpcd_write_payload(struct drm_dp_mst_topology_mgr *mgr, return ret; } +static int do_get_act_status(struct drm_dp_aux *aux) +{ + int ret; + u8 status; + + ret = drm_dp_dpcd_readb(aux, DP_PAYLOAD_TABLE_UPDATE_STATUS, &status); + if (ret < 0) + return ret; + + return status; +} /** * drm_dp_check_act_status() - Polls for ACT handled status. * @mgr: manager to use * * Tries waiting for the MST hub to finish updating it's payload table by - * polling for the ACT handled bit. + * polling for the ACT handled bit for up to 3 seconds (yes-some hubs really + * take that long). * * Returns: * 0 if the ACT was handled in time, negative error code on failure. */ int drm_dp_check_act_status(struct drm_dp_mst_topology_mgr *mgr) { - int count = 0, ret; - u8 status; - - do { - ret = drm_dp_dpcd_readb(mgr->aux, - DP_PAYLOAD_TABLE_UPDATE_STATUS, - &status); - if (ret < 0) { - DRM_DEBUG_KMS("failed to read payload table status %d\n", - ret); - return ret; - } - - if (status & DP_PAYLOAD_ACT_HANDLED) - break; - count++; - udelay(100); - } while (count < 30); - - if (!(status & DP_PAYLOAD_ACT_HANDLED)) { - DRM_DEBUG_KMS("failed to get ACT bit %d after %d retries\n", - status, count); + /* + * There doesn't seem to be any recommended retry count or timeout in + * the MST specification. Since some hubs have been observed to take + * over 1 second to update their payload allocations under certain + * conditions, we use a rather large timeout value. + */ + const int timeout_ms = 3000; + int ret, status; + + ret = readx_poll_timeout(do_get_act_status, mgr->aux, status, + status & DP_PAYLOAD_ACT_HANDLED || status < 0, + 100, timeout_ms * USEC_PER_MSEC); + if (ret < 0 && status >= 0) { + DRM_DEBUG_KMS("Failed to get ACT bit %d after %dms\n", + status, timeout_ms); return -EINVAL; + } else if (status < 0) { + DRM_DEBUG_KMS("Failed to read payload table status: %d\n", + status); + return status; } + return 0; } EXPORT_SYMBOL(drm_dp_check_act_status);
Currently we only poll for an ACT up to 30 times, with a busy-wait delay of 100µs between each attempt - giving us a timeout of 2900µs. While this might seem sensible, it would appear that in certain scenarios it can take dramatically longer then that for us to receive an ACT. On one of the EVGA MST hubs that I have available, I observed said hub sometimes taking longer then a second before signalling the ACT. These delays mostly seem to occur when previous sideband messages we've sent are NAKd by the hub, however it wouldn't be particularly surprising if it's possible to reproduce times like this simply by introducing branch devices with large LCTs since payload allocations have to take effect on every downstream device up to the payload's target. So, instead of just retrying 30 times we poll for the ACT for up to 3ms, and additionally use usleep_range() to avoid a very long and rude busy-wait. Note that the previous retry count of 30 appears to have been arbitrarily chosen, as I can't find any mention of a recommended timeout or retry count for ACTs in the DisplayPort 2.0 specification. This also goes for the range we were previously using for udelay(), although I suspect that was just copied from the recommended delay for link training on SST devices. Changes since v1: * Use readx_poll_timeout() instead of open-coding timeout loop - Sean Paul Signed-off-by: Lyude Paul <lyude@redhat.com> Fixes: ad7f8a1f9ced ("drm/helper: add Displayport multi-stream helper (v0.6)") Cc: Sean Paul <sean@poorly.run> Cc: <stable@vger.kernel.org> # v3.17+ --- drivers/gpu/drm/drm_dp_mst_topology.c | 57 ++++++++++++++++----------- 1 file changed, 34 insertions(+), 23 deletions(-)