Hi!
I am working on trying t=
o figure out how an Android phone's display works by modifying the orig=
inal driver, so I can understand how it works to write a less cluttered, an=
d more non-Android distro friendly driver.
What I =
noticed is that a register needs to be set in order for the screen to updat=
e (I guess there's some kind of internal double buffering going on in t=
he background). I couldn't find a way for it to update automatically, s=
o for now I included the code that sets the register in a VSync thread crea=
ted originally to let the android HAL know that a VSync IRQ has occurred or=
a timeout has passed.
I feel like this isn=
9;t what the solution should be. While looking around I found the fb_sync f=
unction. I thought that sounds like that refresh register should be set the=
re, so I moved the code in fb_sync and I added it to the fb_ops struct. Now=
the display only refreshes on the first frame of the fbcon, replacing the =
phone's bootloader logo with the console. What am I doing wrong? How is=
this supposed to be handled?
Btw I am using a 4.4=
kernel fork, since that's the only kernel that ever shipped with the p=
hone or the SoC, and there's no mainline support for it yet.
<=
/div>
--000000000000f688fc05c3816a0a--
--===============2847503912325621561==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
_______________________________________________
Kernelnewbies mailing list
Kernelnewbies-at-kernelnewbies.org
https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
--===============2847503912325621561==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
_______________________________________________
Hangout mailing list
Hangout-at-nylxs.com
http://lists.mrbrklyn.com/mailman/listinfo/hangout
--===============2847503912325621561==--
--===============2847503912325621561==
Content-Type: multipart/alternative; boundary="000000000000f688fc05c3816a0a"
--000000000000f688fc05c3816a0a
Content-Type: text/plain; charset="UTF-8"
Hi!
I am working on trying to figure out how an Android phone's display works
by modifying the original driver, so I can understand how it works to write
a less cluttered, and more non-Android distro friendly driver.
What I noticed is that a register needs to be set in order for the screen
to update (I guess there's some kind of internal double buffering going on
in the background). I couldn't find a way for it to update automatically,
so for now I included the code that sets the register in a VSync thread
created originally to let the android HAL know that a VSync IRQ has
occurred or a timeout has passed.
I feel like this isn't what the solution should be. While looking around I
found the fb_sync function. I thought that sounds like that refresh
register should be set there, so I moved the code in fb_sync and I added it
to the fb_ops struct. Now the display only refreshes on the first frame of
the fbcon, replacing the phone's bootloader logo with the console. What am
I doing wrong? How is this supposed to be handled?
Btw I am using a 4.4 kernel fork, since that's the only kernel that ever
shipped with the phone or the SoC, and there's no mainline support for it
yet.
--000000000000f688fc05c3816a0a
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Hi!
I am working on trying t=
o figure out how an Android phone's display works by modifying the orig=
inal driver, so I can understand how it works to write a less cluttered, an=
d more non-Android distro friendly driver.
What I =
noticed is that a register needs to be set in order for the screen to updat=
e (I guess there's some kind of internal double buffering going on in t=
he background). I couldn't find a way for it to update automatically, s=
o for now I included the code that sets the register in a VSync thread crea=
ted originally to let the android HAL know that a VSync IRQ has occurred or=
a timeout has passed.
I feel like this isn=
9;t what the solution should be. While looking around I found the fb_sync f=
unction. I thought that sounds like that refresh register should be set the=
re, so I moved the code in fb_sync and I added it to the fb_ops struct. Now=
the display only refreshes on the first frame of the fbcon, replacing the =
phone's bootloader logo with the console. What am I doing wrong? How is=
this supposed to be handled?
Btw I am using a 4.4=
kernel fork, since that's the only kernel that ever shipped with the p=
hone or the SoC, and there's no mainline support for it yet.
<=
/div>
--000000000000f688fc05c3816a0a--
--===============2847503912325621561==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
_______________________________________________
Kernelnewbies mailing list
Kernelnewbies-at-kernelnewbies.org
https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
--===============2847503912325621561==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
_______________________________________________
Hangout mailing list
Hangout-at-nylxs.com
http://lists.mrbrklyn.com/mailman/listinfo/hangout
--===============2847503912325621561==--
