diff mbox series

drm/dp_mst: Correct the bug in drm_dp_update_payload_part1()

Message ID 20191202035841.16984-1-Wayne.Lin@amd.com (mailing list archive)
State New, archived
Headers show
Series drm/dp_mst: Correct the bug in drm_dp_update_payload_part1() | expand

Commit Message

Wayne Lin Dec. 2, 2019, 3:58 a.m. UTC
[Why]
If the payload_state is DP_PAYLOAD_DELETE_LOCAL in series, current
code doesn't delete the payload at current index and just move the
index to next one after shuffling payloads.

[How]
After shuffling payloads, decide whether to move on index or not
according to payload_state of current payload.

Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
---
 drivers/gpu/drm/drm_dp_mst_topology.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Comments

Liu, Zhan Dec. 2, 2019, 4:03 p.m. UTC | #1
> -----Original Message-----
> From: amd-gfx <amd-gfx-bounces@lists.freedesktop.org> On Behalf Of
> Wayne Lin
> Sent: 2019/December/01, Sunday 10:59 PM
> To: dri-devel@lists.freedesktop.org; amd-gfx@lists.freedesktop.org
> Cc: Zuo, Jerry <Jerry.Zuo@amd.com>; Wentland, Harry
> <Harry.Wentland@amd.com>; Kazlauskas, Nicholas
> <Nicholas.Kazlauskas@amd.com>; Lin, Wayne <Wayne.Lin@amd.com>
> Subject: [PATCH] drm/dp_mst: Correct the bug in
> drm_dp_update_payload_part1()
> 
> [Why]
> If the payload_state is DP_PAYLOAD_DELETE_LOCAL in series, current code
> doesn't delete the payload at current index and just move the index to next
> one after shuffling payloads.
> 
> [How]
> After shuffling payloads, decide whether to move on index or not according
> to payload_state of current payload.
> 
> Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>


Reviewed-by: Zhan Liu <zhan.liu@amd.com>


> ---
>  drivers/gpu/drm/drm_dp_mst_topology.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c
> b/drivers/gpu/drm/drm_dp_mst_topology.c
> index 81e92b260d7a..8da5d461ea01 100644
> --- a/drivers/gpu/drm/drm_dp_mst_topology.c
> +++ b/drivers/gpu/drm/drm_dp_mst_topology.c
> @@ -3176,7 +3176,8 @@ int drm_dp_update_payload_part1(struct
> drm_dp_mst_topology_mgr *mgr)
>  			drm_dp_mst_topology_put_port(port);
>  	}
> 
> -	for (i = 0; i < mgr->max_payloads; i++) {
> +	for (i = 0; i < mgr->max_payloads;
> +		(mgr->payloads[i].payload_state ==
> DP_PAYLOAD_DELETE_LOCAL) ? i :
> +i++) {
>  		if (mgr->payloads[i].payload_state !=
> DP_PAYLOAD_DELETE_LOCAL)
>  			continue;
> 
> --
> 2.17.1
> 
> _______________________________________________
> amd-gfx mailing list
> amd-gfx@lists.freedesktop.org
> https://nam11.safelinks.protection.outlook.com/?url=https%3A%2F%2Flists.
> freedesktop.org%2Fmailman%2Flistinfo%2Famd-
> gfx&amp;data=02%7C01%7Czhan.liu%40amd.com%7Cacbf6c9272c44e5642
> ca08d776dbf688%7C3dd8961fe4884e608e11a82d994e183d%7C0%7C0%7C6
> 37108559483579263&amp;sdata=JNUpYWZRxNe%2B0%2FCe04fjWRvWh%2
> FPp3CASy4FRrg0iHuI%3D&amp;reserved=0
Lyude Paul Dec. 3, 2019, 12:23 a.m. UTC | #2
On Mon, 2019-12-02 at 11:58 +0800, Wayne Lin wrote:
> [Why]
> If the payload_state is DP_PAYLOAD_DELETE_LOCAL in series, current
> code doesn't delete the payload at current index and just move the
> index to next one after shuffling payloads.
> 
> [How]
> After shuffling payloads, decide whether to move on index or not
> according to payload_state of current payload.
> 
> Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
> ---
>  drivers/gpu/drm/drm_dp_mst_topology.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c
> b/drivers/gpu/drm/drm_dp_mst_topology.c
> index 81e92b260d7a..8da5d461ea01 100644
> --- a/drivers/gpu/drm/drm_dp_mst_topology.c
> +++ b/drivers/gpu/drm/drm_dp_mst_topology.c
> @@ -3176,7 +3176,8 @@ int drm_dp_update_payload_part1(struct
> drm_dp_mst_topology_mgr *mgr)
>  			drm_dp_mst_topology_put_port(port);
>  	}
>  
> -	for (i = 0; i < mgr->max_payloads; i++) {
> +	for (i = 0; i < mgr->max_payloads;
> +		(mgr->payloads[i].payload_state == DP_PAYLOAD_DELETE_LOCAL) ?
> i : i++) {

Took me a moment to figure out what this line was actually doing. Nice catch
by the way!

Anyway: let's just drop this line to avoid making things confusing to read,
drop i++ from the for loop instead, and just rewrite it so it looks like this:

for (i = 0; i < mgr->max_payloads; /* do nothing */) {
	if (mgr->payloads[i].payload_state != DP_PAYLOAD_DELETE_LOCAL) {
		i++;
		continue;
	}

With those changes, this patch is:

Reviewed-by: Lyude Paul <lyude@redhat.com>

I can go ahead and push these patches to drm-misc for you once you've sent me
the v2
>  		if (mgr->payloads[i].payload_state != DP_PAYLOAD_DELETE_LOCAL)
>  			continue;
>
Wayne Lin Dec. 3, 2019, 4:18 a.m. UTC | #3
> -----Original Message-----
> From: Lyude Paul <lyude@redhat.com>
> Sent: Tuesday, December 3, 2019 8:23 AM
> To: Lin, Wayne <Wayne.Lin@amd.com>; dri-devel@lists.freedesktop.org;
> amd-gfx@lists.freedesktop.org
> Cc: Kazlauskas, Nicholas <Nicholas.Kazlauskas@amd.com>; Wentland, Harry
> <Harry.Wentland@amd.com>; Zuo, Jerry <Jerry.Zuo@amd.com>
> Subject: Re: [PATCH] drm/dp_mst: Correct the bug in
> drm_dp_update_payload_part1()
> 
> On Mon, 2019-12-02 at 11:58 +0800, Wayne Lin wrote:
> > [Why]
> > If the payload_state is DP_PAYLOAD_DELETE_LOCAL in series, current
> > code doesn't delete the payload at current index and just move the
> > index to next one after shuffling payloads.
> >
> > [How]
> > After shuffling payloads, decide whether to move on index or not
> > according to payload_state of current payload.
> >
> > Signed-off-by: Wayne Lin <Wayne.Lin@amd.com>
> > ---
> >  drivers/gpu/drm/drm_dp_mst_topology.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c
> > b/drivers/gpu/drm/drm_dp_mst_topology.c
> > index 81e92b260d7a..8da5d461ea01 100644
> > --- a/drivers/gpu/drm/drm_dp_mst_topology.c
> > +++ b/drivers/gpu/drm/drm_dp_mst_topology.c
> > @@ -3176,7 +3176,8 @@ int drm_dp_update_payload_part1(struct
> > drm_dp_mst_topology_mgr *mgr)
> >  			drm_dp_mst_topology_put_port(port);
> >  	}
> >
> > -	for (i = 0; i < mgr->max_payloads; i++) {
> > +	for (i = 0; i < mgr->max_payloads;
> > +		(mgr->payloads[i].payload_state == DP_PAYLOAD_DELETE_LOCAL) ?
> > i : i++) {
> 
> Took me a moment to figure out what this line was actually doing. Nice catch
> by the way!
> 
> Anyway: let's just drop this line to avoid making things confusing to read, drop
> i++ from the for loop instead, and just rewrite it so it looks like this:
> 
> for (i = 0; i < mgr->max_payloads; /* do nothing */) {
> 	if (mgr->payloads[i].payload_state != DP_PAYLOAD_DELETE_LOCAL) {
> 		i++;
> 		continue;
> 	}
> 
> With those changes, this patch is:
> 
> Reviewed-by: Lyude Paul <lyude@redhat.com>
> 

Thanks for your time and sorry for not well organized code.
I will send you the v2 right away. Thanks!
 
> I can go ahead and push these patches to drm-misc for you once you've sent
> me the v2
> >  		if (mgr->payloads[i].payload_state != DP_PAYLOAD_DELETE_LOCAL)
> >  			continue;
> >
> --
> Cheers,
> 	Lyude Paul
--
BR,
Wayne Lin
diff mbox series

Patch

diff --git a/drivers/gpu/drm/drm_dp_mst_topology.c b/drivers/gpu/drm/drm_dp_mst_topology.c
index 81e92b260d7a..8da5d461ea01 100644
--- a/drivers/gpu/drm/drm_dp_mst_topology.c
+++ b/drivers/gpu/drm/drm_dp_mst_topology.c
@@ -3176,7 +3176,8 @@  int drm_dp_update_payload_part1(struct drm_dp_mst_topology_mgr *mgr)
 			drm_dp_mst_topology_put_port(port);
 	}
 
-	for (i = 0; i < mgr->max_payloads; i++) {
+	for (i = 0; i < mgr->max_payloads;
+		(mgr->payloads[i].payload_state == DP_PAYLOAD_DELETE_LOCAL) ? i : i++) {
 		if (mgr->payloads[i].payload_state != DP_PAYLOAD_DELETE_LOCAL)
 			continue;