Bug#980536: CMA=64M insufficient for starting firefox on RPi4B with 2xFullHD

2023-10-13 Thread ibu radempa

Hi Gunnar,

on a RPi4B with 8G RAM and two Full HD displays connected I am running 
bookworm. After upgrading firefox (102.15.1esr-1~deb12u1 to 
115.3.0esr-1~deb12u1) recently, it failed to start and threw multiply:


```
DRM_IOCTL_MODE_CREATE_DUMB failed: Cannot allocate memory
Failed to create scanout resource
```

After setting CMA=128M in /etc/default/raspi-firmware, update-initramfs 
and reboot firefox started properly and is usable.


```
# grep -i cma /proc/meminfo
CmaTotal: 131072 kB
CmaFree:   48612 kB
```

And if I run mpv (with CMA=128M) in addition to firefox, I occasionally 
also get the above error.


I suggest removing the notice in /etc/default/raspi-firmware concerning 
RPi4, or referring to this bug, and maybe updating other docs and the 
image when you find time, because I guess firefox is a frequently used 
application on this device and if it fails, it would push off users.


Cheers,
ibu



Bug#969287: raspi-firmware: Dual HDMI support for RaspberryPI4

2021-06-25 Thread ibu radempa

On 25.06.21 18:52, Diederik de Haas wrote:

On Wed, 6 Jan 2021 14:11:02 -0600 Gunnar Wolf  wrote:

So, do you still have this behavior? Have you (or has anybody else)
been able to get dual-screen support on your RPi4? Or do you have any
promising pointers on this?


Making sure the submitter receives the the reply/request for info.


Indeed, I missed Gunnars message in January. Now I tried immediately and 
in fact, dual HDMI works (using bullseye's 5.10.0-6-arm64 kernel).

Disconnecting and reconnecting the second display also worked fine.

Bug solved!

Greetings from a very happy user :)
ibu



Bug#909498: duplicate of #908449

2018-09-27 Thread ibu radempa
Hello Anthony,

today I also encountered a crash of firefox-esr 60.2.1esr-1~deb9u1 on
armhf from stable-updates. After installing 60.2.1esr-1 from unstable
the problem was gone.

Hello Maintainer,

this bug obviously is a duplicate of #908449 and has been fixed in unstable.

Thanks,
ibu



Bug#593847: conflict with rhythmbox and gvfsd

2012-03-25 Thread ibu radempa
On squeeze with

gphoto 2.4.5-3
libgphoto2-2 2.4.6-3
libgphoto2-port0 2.4.6-3
rhythmbox 0.12.8-3

after some testing I ran lsof on the USB device file
(/dev/bus/usb/003/025 in my case) and found that it was also claimed by
rhythmbox (an application which is usually started upon login). After
stopping rhythmbox I could read data from the camera.

After unplugging and replugging the device was claimed by gvfsd, so I
ran gvfs -s gphoto2 and again could read data from it.



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org