Revert "mei: hdcp: Replace one-element array with flexible-array member"
Subsystems (2)
Greg please revert, this commit it changes size of struct wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit c56967d674e361ebe716e66992e3c5332b25ac1f.
--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)
drivers/misc/mei/hdcp/mei_hdcp.c +1 -1
@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type ))); verify_mprime_in.k =cpu_to_be16 (data->k);
drivers/misc/mei/hdcp/mei_hdcp.h +1 -1
@@ -358,7 +358,7 @@ struct wired_cmd_repeater_auth_stream_req_in {u8 seq_num_m[HDCP_2_2_SEQ_NUM_LEN ];u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Signature · 1 line
2.25.4
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.
Could you elaborate on what's the firmware expecting, exactly?
Thanks
--
Gustavo
Quoted 40 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +1
-1@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type ))); verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +1
-1@@ -358,7 +358,7 @@ struct wired_cmd_repeater_auth_stream_req_in {u8 seq_num_m[HDCP_2_2_SEQ_NUM_LEN ];u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 12 lines
-----Original Message-----
From: Gustavo A. R. Silva <[email protected]>
Sent: Wednesday, July 29, 2020 00:57
To: Winkler, Tomas <[email protected]>; Greg Kroah-Hartman <[email protected]>
Cc: Usyskin, Alexander <[email protected]>; linux- [email protected]; Gustavo A . R . Silva <[email protected]>
Subject: Re: [char-misc-next] Revert "mei: hdcp: Replace one-element array with flexible-array member"
Quoted 8 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
struct wired_cmd_repeater_auth_stream_req_in {
- struct hdcp2_streamid_type streams[1];
+ struct hdcp2_streamid_type streams[];
}
But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is MAINTAINERS file for a reason.
Quoted 50 lines
Thanks
--
Gustavodrivers/misc/mei/hdcp/mei_hdcp.c static int mei_hdcp_verify_mprime(struct device
Quoted 19 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(struct device*dev,
Quoted 15 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data- seq_num_m);memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type ))); verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0drivers/misc/mei/hdcp/mei_hdcp.h struct wired_cmd_repeater_auth_stream_req_in { · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct wired_cmd_repeater_auth_stream_req_in { u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 20 lines · Tomas
Quoted 12 lines
-----Original Message-----
From: Gustavo A. R. Silva <[email protected]>
Sent: Wednesday, July 29, 2020 00:57
To: Winkler, Tomas <[email protected]>; Greg Kroah-Hartman <[email protected]>
Cc: Usyskin, Alexander <[email protected]>; linux- [email protected]; Gustavo A . R . Silva <[email protected]>
Subject: Re: [char-misc-next] Revert "mei: hdcp: Replace one-element array with flexible-array member"Quoted 8 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
Quoted 9 lines
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));
I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.
I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback
and this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) [email protected] (open list)
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and lists?
Thanks
--
Gustavo
Quoted 46 lines
Quoted 46 lines
drivers/misc/mei/hdcp/mei_hdcp.c static int mei_hdcp_verify_mprime(struct device
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(struct device*dev,
Quoted 15 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data- seq_num_m);memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type ))); verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0drivers/misc/mei/hdcp/mei_hdcp.h struct wired_cmd_repeater_auth_stream_req_in { · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct wired_cmd_repeater_auth_stream_req_in { u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 56 lines · Gustavo A. R. Silva
Quoted 20 lines · Tomas
Quoted 12 lines
-----Original Message-----
From: Gustavo A. R. Silva <[email protected]>
Sent: Wednesday, July 29, 2020 00:57
To: Winkler, Tomas <[email protected]>; Greg Kroah-Hartman <[email protected]>
Cc: Usyskin, Alexander <[email protected]>; linux- [email protected]; Gustavo A . R . Silva <[email protected]>
Subject: Re: [char-misc-next] Revert "mei: hdcp: Replace one-element array with flexible-array member"Quoted 8 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
Quoted 9 lines
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) [email protected] (open list)
Quoted 2 lines
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and lists?>
Notice that if I run this command:
$ scripts/get_maintainer.pl -f drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) Ramalingam C <[email protected]> (commit_signer:2/2=100%,authored:2/2=100%,added_lines:4/4=100%,removed_lines:13/13=100%) Tomas Winkler <[email protected]> (commit_signer:2/2=100%) Uma Shankar <[email protected]> (commit_signer:2/2=100%) Jani Nikula <[email protected]> (commit_signer:2/2=100%) Shashank Sharma <[email protected]> (commit_signer:2/2=100%) [email protected] (open list)
You appear this time, but not as a maintainer.
Thanks
--
Gustavo
Quoted 46 lines
Quoted 46 lines
Quoted 46 lines
drivers/misc/mei/hdcp/mei_hdcp.c static int mei_hdcp_verify_mprime(struct device
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(struct device*dev,
Quoted 15 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data- seq_num_m);memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type ))); verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0drivers/misc/mei/hdcp/mei_hdcp.h struct wired_cmd_repeater_auth_stream_req_in { · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct wired_cmd_repeater_auth_stream_req_in { u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 70 lines
-----Original Message-----
From: Gustavo A. R. Silva <[email protected]>
Sent: Wednesday, July 29, 2020 01:15
To: Winkler, Tomas <[email protected]>; Greg Kroah-Hartman <[email protected]>
Cc: Usyskin, Alexander <[email protected]>; linux- [email protected]; Gustavo A . R . Silva <[email protected]>
Subject: Re: [char-misc-next] Revert "mei: hdcp: Replace one-element array with flexible-array member"Quoted 20 lines · Tomas
Quoted 12 lines
-----Original Message-----
From: Gustavo A. R. Silva <[email protected]>
Sent: Wednesday, July 29, 2020 00:57
To: Winkler, Tomas <[email protected]>; Greg Kroah-Hartman <[email protected]>
Cc: Usyskin, Alexander <[email protected]>; linux- [email protected]; Gustavo A . R . Silva <[email protected]>
Subject: Re: [char-misc-next] Revert "mei: hdcp: Replace one-element array with flexible-array member"Quoted 8 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
Quoted 9 lines
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list)
Quoted 2 lines
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and lists?
Not sure, it should be caught by drivers/misc/mei/* ? Maybe it is not recursive? Need to check the script, frankly I usually check this manually.
INTEL MANAGEMENT ENGINE (mei)
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/* F: drivers/watchdog/mei_wdt.c F: include/linux/mei_cl_bus.h F: include/uapi/linux/mei.h F: samples/mei/*
Quoted 52 lines
Thanks
--
GustavoQuoted 21 lines
Quoted 21 lines
drivers/misc/mei/hdcp/mei_hdcp.c static int mei_hdcp_verify_mprime(struct device
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(struct device*dev,
Quoted 2 lines
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c static int mei_hdcp_verify_mprime(struct device · line 573
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0HDCP_2_2_MPRIME_LEN); drm_hdcp_cpu_to_be24(verify_mprime_in.seq_num_m, data-
seq_num_m);Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type )));Quoted 8 lines
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct wired_cmd_repeater_auth_stream_req_in
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct wired_cmd_repeater_auth_stream_req_in{
Quoted 10 lines
Quoted 10 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct wired_cmd_repeater_auth_stream_req_in · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 54 lines
Quoted 45 lines
Quoted 9 lines
Quoted 9 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
Quoted 9 lines
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list)Quoted 2 lines
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and
lists?
Not sure, it should be caught by
drivers/misc/mei/*?
Maybe it is not recursive? Need to check the script, frankly I usually check this manually.INTEL MANAGEMENT ENGINE (mei)
Quoted 10 lines
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/* F: drivers/watchdog/mei_wdt.c F: include/linux/mei_cl_bus.h F: include/uapi/linux/mei.h F: samples/mei/*
It's not recursive, when I add drivers/misc/mei/hdcp/* . It works correctly I will post a patch.
Thanks
Tomas
Quoted 55 lines
Quoted 35 lines
Thanks
--
GustavoQuoted 22 lines
Quoted 22 lines
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(structQuoted 1 line
device
*dev,
Quoted 2 lines
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0HDCP_2_2_MPRIME_LEN); drm_hdcp_cpu_to_be24(verify_mprime_in.seq_num_m, data-
seq_num_m);Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type )));Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct
wired_cmd_repeater_auth_stream_req_inQuoted 11 lines
{
Quoted 10 lines
Quoted 10 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Tomas,
Please, see some comments below...
Quoted 14 lines · Tomas
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
Quoted 9 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));
I don't think the fix for this is to add 1 byte, if any, it seems to be sizeof(*data->streams), or sizeof(struct hdcp2_streamid_type) what needs to be added.
But it might be better to code something like this, instead:
drivers/misc/mei/hdcp/mei_hdcp.c +2 -2
@@ -572,12 +572,12 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, - (data->k *sizeof (struct hdcp2_streamid_type ))); + (data->k *sizeof (*data->streams))); verify_mprime_in.k =cpu_to_be16 (data->k); byte =mei_cldev_send (cldev, (u8 *)&verify_mprime_in, -sizeof (verify_mprime_in)); +struct_size (&verify_mprime_in, streams, data->k));if (byte <0 ) {dev_dbg (dev,"mei_cldev_send failed. %zd\n" , byte);return byte;
struct_size(&verify_mprime_in, streams, data->k) will give us the size, in bytes, of struct wired_cmd_repeater_auth_stream_req_in plus the size in bytes for the streams[] flexible-array, which is determined by struct hdcp2_streamid_type and data->k.
What do you think?
See more comments below...
Quoted 47 lines
Quoted 36 lines
Quoted 27 lines
I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list)Quoted 2 lines
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and
lists?
Not sure, it should be caught by
drivers/misc/mei/*?
Maybe it is not recursive? Need to check the script, frankly I usually check this manually.INTEL MANAGEMENT ENGINE (mei)
Quoted 10 lines
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/* F: drivers/watchdog/mei_wdt.c F: include/linux/mei_cl_bus.h F: include/uapi/linux/mei.h F: samples/mei/*
Quoted 3 lines
It's not recursive, when I add
drivers/misc/mei/hdcp/*. It works correctly I will post a patch.
Great. I'm glad we got to the root cause of this issue. :)
Thanks
--
Gustavo
Quoted 59 lines
Thanks
TomasQuoted 55 lines
Quoted 35 lines
Thanks
--
GustavoQuoted 22 lines
Quoted 22 lines
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(structQuoted 1 line
device
*dev,
Quoted 2 lines
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0HDCP_2_2_MPRIME_LEN); drm_hdcp_cpu_to_be24(verify_mprime_in.seq_num_m, data-
seq_num_m);Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type )));Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct
wired_cmd_repeater_auth_stream_req_inQuoted 11 lines
{
Quoted 10 lines
Quoted 10 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 55 lines
Tomas,
Please, see some comments below...
Quoted 14 lines · Tomas
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
Quoted 9 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));I don't think the fix for this is to add 1 byte, if any, it seems to be sizeof(*data-
Quoted 1 line
streams), or
sizeof(struct hdcp2_streamid_type)what needs to be added.But it might be better to code something like this, instead:
drivers/misc/mei/hdcp/mei_hdcp.c +3
-2@@ -572,12 +572,12 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data- >seq_num_m);memcpy (verify_mprime_in.streams, data->streams, - (data->k *sizeof (struct hdcp2_streamid_type ))); + (data->k *sizeof (*data->streams))); verify_mprime_in.k =cpu_to_be16 (data->k); byte =mei_cldev_send (cldev, (u8 *)&verify_mprime_in, -sizeof (verify_mprime_in)); +struct_size (&verify_mprime_in, streams, + data->k));if (byte <0 ) {dev_dbg (dev,"mei_cldev_send failed. %zd\n" , byte);return byte;
Quoted 7 lines
struct_size(&verify_mprime_in, streams, data->k)will give us the size, in bytes, of structwired_cmd_repeater_auth_stream_req_inplus the size in bytes for the streams[] flexible-array, which is determined by structhdcp2_streamid_typeanddata->k.What do you think?
I think I will talk to the firmware guys tomorrow morning...
Quoted 121 lines
See more comments below...
Quoted 48 lines
Quoted 37 lines
Quoted 27 lines
I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list)Quoted 3 lines
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and
lists?
Not sure, it should be caught by
drivers/misc/mei/*? Maybe it is not recursive? Need to check the script, frankly I usually check this manually.INTEL MANAGEMENT ENGINE (mei)
Quoted 10 lines
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/* F: drivers/watchdog/mei_wdt.c F: include/linux/mei_cl_bus.h F: include/uapi/linux/mei.h F: samples/mei/*Quoted 3 lines
It's not recursive, when I add
drivers/misc/mei/hdcp/*. It works correctly I will post a patch.Great. I'm glad we got to the root cause of this issue. :)
Thanks
--
GustavoQuoted 52 lines
Thanks
TomasQuoted 48 lines
Quoted 35 lines
Thanks
--
GustavoQuoted 22 lines
Quoted 22 lines
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(structQuoted 1 line
device
*dev,
Quoted 2 lines
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0HDCP_2_2_MPRIME_LEN); drm_hdcp_cpu_to_be24(verify_mprime_in.seq_num_m, data-
seq_num_m);Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type )));Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct
wired_cmd_repeater_auth_stream_req_inQuoted 4 lines
{
Quoted 3 lines
Quoted 3 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8m_prime[HDCP_2_2_MPRIME_LEN];Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 50 lines · Gustavo A. R. Silva
Tomas,
Please, see some comments below...
Quoted 14 lines · Tomas
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
Quoted 9 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.Could you elaborate on what's the firmware expecting, exactly?
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));I don't think the fix for this is to add 1 byte, if any, it seems to be
sizeof(*data->streams), orsizeof(struct hdcp2_streamid_type)what needs to be added.But it might be better to code something like this, instead:
drivers/misc/mei/hdcp/mei_hdcp.c +2
-2@@ -572,12 +572,12 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, - (data->k *sizeof (struct hdcp2_streamid_type ))); + (data->k *sizeof (*data->streams))); verify_mprime_in.k =cpu_to_be16 (data->k); byte =mei_cldev_send (cldev, (u8 *)&verify_mprime_in, -sizeof (verify_mprime_in)); +struct_size (&verify_mprime_in, streams, data->k));if (byte <0 ) {dev_dbg (dev,"mei_cldev_send failed. %zd\n" , byte);return byte;
Quoted 7 lines
struct_size(&verify_mprime_in, streams, data->k)will give us the size, in bytes, of structwired_cmd_repeater_auth_stream_req_inplus the size in bytes for the streams[] flexible-array, which is determined by structhdcp2_streamid_typeanddata->k.What do you think?
This is even better:
drivers/misc/mei/hdcp/mei_hdcp.c +2 -2
@@ -572,12 +572,12 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, - (data->k *sizeof (struct hdcp2_streamid_type ))); +flex_array_size (&verify_mprime_in, streams, data->k)); verify_mprime_in.k =cpu_to_be16 (data->k); byte =mei_cldev_send (cldev, (u8 *)&verify_mprime_in, -sizeof (verify_mprime_in)); +struct_size (&verify_mprime_in, streams, data->k));if (byte <0 ) {dev_dbg (dev,"mei_cldev_send failed. %zd\n" , byte);return byte;
the flex_array_size() is a new helper that was designed for this situations. :) It calculates the size of a flexible array member within an enclosing structure, which is exactly the case.
Thanks
--
Gustavo
Quoted 118 lines
See more comments below...
Quoted 47 lines
Quoted 36 lines
Quoted 27 lines
I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list)Quoted 2 lines
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and
lists?
Not sure, it should be caught by
drivers/misc/mei/*?
Maybe it is not recursive? Need to check the script, frankly I usually check this manually.INTEL MANAGEMENT ENGINE (mei)
Quoted 10 lines
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/* F: drivers/watchdog/mei_wdt.c F: include/linux/mei_cl_bus.h F: include/uapi/linux/mei.h F: samples/mei/*Quoted 3 lines
It's not recursive, when I add
drivers/misc/mei/hdcp/*. It works correctly I will post a patch.Great. I'm glad we got to the root cause of this issue. :)
Thanks
--
GustavoQuoted 59 lines
Thanks
TomasQuoted 55 lines
Quoted 35 lines
Thanks
--
GustavoQuoted 22 lines
Quoted 22 lines
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(structQuoted 1 line
device
*dev,
Quoted 2 lines
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0HDCP_2_2_MPRIME_LEN); drm_hdcp_cpu_to_be24(verify_mprime_in.seq_num_m, data-
seq_num_m);Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type )));Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct
wired_cmd_repeater_auth_stream_req_inQuoted 11 lines
{
Quoted 10 lines
Quoted 10 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 9 lines · Gustavo A. R. Silva
Quoted 6 lines · Gustavo A. R. Silva
Tomas,
Please, see some comments below...
Quoted 1 line · Tomas
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Hi Tomas,
Quoted 3 lines · Tomas Winkler
Greg please revert, this commit it changes size of struct
wired_cmd_repeater_auth_stream_req_in, this is not what firmware is expecting.
Instead of reverting that commit, I think this patch should be applied to fix it, if Tomas consider this is the appropiate solution:
drivers/misc/mei/hdcp/mei_hdcp.c +2 -2
@@ -572,12 +572,12 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); +flex_array_size (&verify_mprime_in, streams, data->k)); verify_mprime_in.k =cpu_to_be16 (data->k); byte =mei_cldev_send (cldev, (u8 *)&verify_mprime_in, -sizeof (verify_mprime_in)); +struct_size (&verify_mprime_in, streams, data->k));if (byte <0 ) {dev_dbg (dev,"mei_cldev_send failed. %zd\n" , byte);return byte;
I'll write up a proper patch after getting some feedback.
Thanks
--
Gustavo
Quoted 193 lines
Quoted 37 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 1 line
Could you elaborate on what's the firmware expecting, exactly?
struct
wired_cmd_repeater_auth_stream_req_in{
- structhdcp2_streamid_typestreams[1];
+ structhdcp2_streamid_typestreams[];
}But then you have, which you haven't changed to + 1 byte = mei_cldev_send(cldev, (u8 *)&verify_mprime_in,
sizeof(verify_mprime_in));I don't think the fix for this is to add 1 byte, if any, it seems to be
sizeof(*data->streams), orsizeof(struct hdcp2_streamid_type)what needs to be added.But it might be better to code something like this, instead:
drivers/misc/mei/hdcp/mei_hdcp.c +2
-2@@ -572,12 +572,12 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, - (data->k *sizeof (struct hdcp2_streamid_type ))); + (data->k *sizeof (*data->streams))); verify_mprime_in.k =cpu_to_be16 (data->k); byte =mei_cldev_send (cldev, (u8 *)&verify_mprime_in, -sizeof (verify_mprime_in)); +struct_size (&verify_mprime_in, streams, data->k));if (byte <0 ) {dev_dbg (dev,"mei_cldev_send failed. %zd\n" , byte);return byte;Quoted 7 lines
struct_size(&verify_mprime_in, streams, data->k)will give us the size, in bytes, of structwired_cmd_repeater_auth_stream_req_inplus the size in bytes for the streams[] flexible-array, which is determined by structhdcp2_streamid_typeanddata->k.What do you think?
This is even better:
drivers/misc/mei/hdcp/mei_hdcp.c +2
-2@@ -572,12 +572,12 @@ static int mei_hdcp_verify_mprime(struct device *dev, HDCP_2_2_MPRIME_LEN);drm_hdcp_cpu_to_be24 (verify_mprime_in.seq_num_m, data->seq_num_m);memcpy (verify_mprime_in.streams, data->streams, - (data->k *sizeof (struct hdcp2_streamid_type ))); +flex_array_size (&verify_mprime_in, streams, data->k)); verify_mprime_in.k =cpu_to_be16 (data->k); byte =mei_cldev_send (cldev, (u8 *)&verify_mprime_in, -sizeof (verify_mprime_in)); +struct_size (&verify_mprime_in, streams, data->k));if (byte <0 ) {dev_dbg (dev,"mei_cldev_send failed. %zd\n" , byte);return byte;the
flex_array_size()is a new helper that was designed for this situations. :) It calculates the size of a flexible array member within an enclosing structure, which is exactly the case.Thanks
--
GustavoQuoted 118 lines
See more comments below...
Quoted 47 lines
Quoted 36 lines
Quoted 27 lines
I see, this is the kind of feedback I need from people that knows the code better. Thanks!
Quoted 4 lines
But that's not the major point. Point is that we should be able to review and test the code before it is merged. You haven't run it, right? There is
MAINTAINERSfile for a reason.I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list)Quoted 2 lines
For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and
lists?
Not sure, it should be caught by
drivers/misc/mei/*?
Maybe it is not recursive? Need to check the script, frankly I usually check this manually.INTEL MANAGEMENT ENGINE (mei)
Quoted 10 lines
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/* F: drivers/watchdog/mei_wdt.c F: include/linux/mei_cl_bus.h F: include/uapi/linux/mei.h F: samples/mei/*Quoted 3 lines
It's not recursive, when I add
drivers/misc/mei/hdcp/*. It works correctly I will post a patch.Great. I'm glad we got to the root cause of this issue. :)
Thanks
--
GustavoQuoted 59 lines
Thanks
TomasQuoted 55 lines
Quoted 35 lines
Thanks
--
GustavoQuoted 22 lines
Quoted 22 lines
Quoted 20 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(structQuoted 1 line
device
*dev,
Quoted 2 lines
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0HDCP_2_2_MPRIME_LEN); drm_hdcp_cpu_to_be24(verify_mprime_in.seq_num_m, data-
seq_num_m);Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type )));Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct
wired_cmd_repeater_auth_stream_req_inQuoted 11 lines
{
Quoted 10 lines
Quoted 10 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 2 lines · Tomas
Quoted 2 lines
-----Original Message-----
From: Gustavo A. R. Silva <[email protected]>
[]
Quoted 18 lines
Quoted 6 lines
I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
Quoted 11 lines
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) For some reason you don't appear on the list above. Do you see anything wrong with the command I'm running to get the maintainers and lists?
Quoted 1 line
Not sure, it should be caught by
drivers/misc/mei/*?
No, this pattern matches _only_ files in that
particular directory and not any files in any
subdirectory.
Quoted 1 line
Maybe it is not recursive?
It depends on the pattern.
Patterns with trailing slashes match subdirectories too.
from MAINTAINERS:
F: *Files* and directories wildcard patterns.
A trailing slash includes all files and subdirectory files.
F: drivers/net/ all files in and below drivers/net
F: drivers/net/* all files in drivers/net, but not below
F: */net/* all files in "any top level directory"/net
One pattern per line. Multiple F: lines acceptable.
Quoted 9 lines
Need to check the script,
frankly I usually check this manually. probINTEL MANAGEMENT ENGINE (mei)
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/*
You probably want:
F: drivers/misc/mei/
Quoted 56 lines
F: drivers/watchdog/mei_wdt.c F: include/linux/mei_cl_bus.h F: include/uapi/linux/mei.h F: samples/mei/*Quoted 50 lines
Thanks
--
GustavoQuoted 20 lines
Quoted 20 lines
drivers/misc/mei/hdcp/mei_hdcp.c static int mei_hdcp_verify_mprime(struct device
Quoted 19 lines
I really do not appreciate that the code is bypassing driver maintaner review, I think this is a minimum we can ask for, this is not for a first time.
This reverts commit
c56967d674e361ebe716e66992e3c5332b25ac1f.Cc Gustavo A. R. Silva <[email protected]>Signed-off-by Tomas Winkler <[email protected]>--- drivers/misc/mei/hdcp/mei_hdcp.c | 2 +- drivers/misc/mei/hdcp/mei_hdcp.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-)drivers/misc/mei/hdcp/mei_hdcp.c +0
-0@@ -572,7 +572,7 @@ static int mei_hdcp_verify_mprime(struct device*dev,
Quoted 2 lines
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c static int mei_hdcp_verify_mprime(struct device · line 573
Quoted 2 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0HDCP_2_2_MPRIME_LEN); drm_hdcp_cpu_to_be24(verify_mprime_in.seq_num_m, data-
seq_num_m);Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
drivers/misc/mei/hdcp/mei_hdcp.c +1
-1memcpy (verify_mprime_in.streams, data->streams, -array_size (data->k,sizeof (*data->streams))); + (data->k *sizeof (struct hdcp2_streamid_type )));Quoted 8 lines
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct wired_cmd_repeater_auth_stream_req_in
Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.c +0
-0verify_mprime_in.k =cpu_to_be16 (data->k);drivers/misc/mei/hdcp/mei_hdcp.h +0
-0@@ -358,7 +358,7 @@ struct wired_cmd_repeater_auth_stream_req_in{
Quoted 10 lines
Quoted 10 lines
drivers/misc/mei/hdcp/mei_hdcp.h struct wired_cmd_repeater_auth_stream_req_in · line 358
Quoted 1 line
drivers/misc/mei/hdcp/mei_hdcp.h +0
-0u8seq_num_m[HDCP_2_2_SEQ_NUM_LEN];Quoted 8 lines
drivers/misc/mei/hdcp/mei_hdcp.h +1
-1u8 m_prime[HDCP_2_2_MPRIME_LEN ];__be16 k; -struct hdcp2_streamid_type streams[]; +struct hdcp2_streamid_type streams[1 ]; } __packed;struct wired_cmd_repeater_auth_stream_req_out {
Quoted 53 lines · Tomas
Quoted 2 lines
Quoted 2 lines
-----Original Message-----
From: Gustavo A. R. Silva <[email protected]>[]
Quoted 6 lines
Quoted 6 lines
I'm using this command:
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallbackand this is the output for both files:
Quoted 11 lines
Quoted 11 lines
$ scripts/get_maintainer.pl --nokeywords --nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.c Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) $ scripts/get_maintainer.pl --nokeywords -- nogit --nogit-fallback drivers/misc/mei/hdcp/mei_hdcp.h Arnd Bergmann <[email protected]> (supporter:CHAR and MISC DRIVERS) Greg Kroah-Hartman <[email protected]> (supporter:CHAR and MISC DRIVERS) linux- [email protected] (open list) For some reason you don't appear on the list above. Do you see anything wrong with the commandI'm running to get the maintainers and lists?
Quoted 2 lines
Not sure, it should be caught by
drivers/misc/mei/*?No, this pattern matches _only_ files in that particular directory and not any files in any subdirectory.
Quoted 1 line
Maybe it is not recursive?
It depends on the pattern.
Patterns with trailing slashes match subdirectories too.from
MAINTAINERS:F: *Files* and directories wildcard patterns.
A trailing slash includes all files and subdirectory files.
F:drivers/net/all files in and belowdrivers/net
F:drivers/net/*all files indrivers/net, but not below
F: */net/* all files in "any top level directory"/net
One pattern per line. Multiple F: lines acceptable.Quoted 9 lines
Need to check the script,
frankly I usually check this manually. probINTEL MANAGEMENT ENGINE (mei)
M: Tomas Winkler <[email protected]> L: [email protected] S: Supported F: Documentation/driver-api/mei/* F: drivers/misc/mei/*
Quoted 3 lines
You probably want:
F:
drivers/misc/mei/
Yes, thanks.
Tomas