On 2021-04-20 16:38, Stephen Boyd wrote:
Quoting Kuogee Hsieh (2021-04-16 10:38:51)
Maybe when the cable is disconnected the DP phy should be shutdown and some bit in the phy could effectively "cut off" the aux channel and then NAKs would start coming through here in the DP controller I/O register space. This patch have DP aux channel read/write to return NAK immediately if DP controller connection status is in unplugged state.
Changes in V3: -- check core_initialized before handle irq_hpd Signed-off-by: Kuogee Hsieh khsieh@codeaurora.org
drivers/gpu/drm/msm/dp/dp_aux.c | 5 +++++ drivers/gpu/drm/msm/dp/dp_display.c | 14 ++++++++++---- drivers/gpu/drm/msm/dp/dp_link.c | 20 +++++++++++++++----- 3 files changed, 30 insertions(+), 9 deletions(-)
diff --git a/drivers/gpu/drm/msm/dp/dp_aux.c b/drivers/gpu/drm/msm/dp/dp_aux.c index 7c22bfe..fae3806 100644 --- a/drivers/gpu/drm/msm/dp/dp_aux.c +++ b/drivers/gpu/drm/msm/dp/dp_aux.c @@ -343,6 +343,11 @@ static ssize_t dp_aux_transfer(struct drm_dp_aux *dp_aux,
mutex_lock(&aux->mutex);
if (!dp_catalog_link_is_connected(aux->catalog)) {ret = -ETIMEDOUT;goto unlock_exit;}This still makes me concerned. Any possibility to not do this and have the phy cut the connection off and have this transfer timeout immediately?
no, we have to wait hardware AUX_NACK timeout. only this or the abort flag used last time. Last time you have kernel crash because of service irq_hpd while clock is turned off. I have add core_initialized checking wiinin irq_hpd to prevent this. I think abort flag approach is safer.
aux->native = msg->request & (DP_AUX_NATIVE_WRITE &DP_AUX_NATIVE_READ);
/* Ignore address only message */diff --git a/drivers/gpu/drm/msm/dp/dp_display.c b/drivers/gpu/drm/msm/dp/dp_display.c index 1784e11..db3f45e 100644 --- a/drivers/gpu/drm/msm/dp/dp_display.c +++ b/drivers/gpu/drm/msm/dp/dp_display.c @@ -571,7 +571,7 @@ static int dp_hpd_plug_handle(struct dp_display_private *dp, u32 data) dp->hpd_state = ST_DISCONNECTED;
if (ret == -ECONNRESET) { /* cable unplugged */
dp->core_initialized = false;
DRM_ERROR("dongle unplugged = %d\n", ret);Is this a debug message?
} } else {@@ -711,9 +711,15 @@ static int dp_irq_hpd_handle(struct dp_display_private *dp, u32 data) return 0; }
ret = dp_display_usbpd_attention_cb(&dp->pdev->dev);if (ret == -ECONNRESET) { /* cable unplugged */dp->core_initialized = false;
/** dp core (ahb/aux clks) must be initialized before* irq_hpd be handled*/if (dp->core_initialized) {ret = dp_display_usbpd_attention_cb(&dp->pdev->dev);if (ret == -ECONNRESET) { /* cable unplugged */DRM_ERROR("dongle unplugged = %d\n", ret);Another debug message?
} } mutex_unlock(&dp->event_mutex);diff --git a/drivers/gpu/drm/msm/dp/dp_link.c b/drivers/gpu/drm/msm/dp/dp_link.c index be986da..53ecae6 100644 --- a/drivers/gpu/drm/msm/dp/dp_link.c +++ b/drivers/gpu/drm/msm/dp/dp_link.c @@ -737,18 +737,25 @@ static int dp_link_parse_sink_count(struct dp_link *dp_link) return 0; }
-static void dp_link_parse_sink_status_field(struct dp_link_private *link) +static int dp_link_parse_sink_status_field(struct dp_link_private *link) { int len = 0;
link->prev_sink_count = link->dp_link.sink_count;
dp_link_parse_sink_count(&link->dp_link);
len = dp_link_parse_sink_count(&link->dp_link);if (len < 0) {DRM_ERROR("DP parse sink count failed\n");return len;} len = drm_dp_dpcd_read_link_status(link->aux, link->link_status);
if (len < DP_LINK_STATUS_SIZE)
if (len < DP_LINK_STATUS_SIZE) { DRM_ERROR("DP link status read failed\n");
dp_link_parse_request(link);
return len;}return dp_link_parse_request(link);}
/** @@ -1032,7 +1039,10 @@ int dp_link_process_request(struct dp_link *dp_link)
dp_link_reset_data(link);
dp_link_parse_sink_status_field(link);
ret = dp_link_parse_sink_status_field(link);if (ret) {return ret;} if (link->request.test_requested == DP_TEST_LINK_EDID_READ) { dp_link->sink_request |= DP_TEST_LINK_EDID_READ;--
Can you split this part off into another patch? It seems to stand on its own as it makes the code more robust to transfer errors in the sink parsing code.