[fedora-arm] "alloc_contig_range: ... PFNs busy" on aarch64

[Toralf Lund]

Hi,

Another slight problem I have with aarch64 Fedora 27 on my Raspberry Pi
3, is that it spews out console messages of the form

alloc_contig_range: ... PFNs busy

Essentially, it's the same behaviour as what's described in
https://bugzilla.redhat.com/show_bug.cgi?id=1387793, but I'm seeing
different address ranges, like [30f00, 31f00), [31000, 32000) and
[31800, 32800). There are some more variants, but this morning, they all
have one of the 3 starting points indicated here.

Questions:

1. Has there been any development related to this issue?
2. Should I file another Bugzilla report - seeing as the original one
is closed with INSUFFICIENT_DATA and belongs to a different release?
3. Is there anything I can do to get around the problem, or at least
suppress the error messages? (They kind of get in my way right now
as I'm experimenting with a framebuffer application.)

Thanks,

- Toralf

[Peter Robinson]

Hi,
Another slight problem I have with aarch64 Fedora 27 on my Raspberry Pi 3,
is that it spews out console messages of the form
alloc_contig_range: ... PFNs busy
Essentially, it's the same behaviour as what's described in
https://bugzilla.redhat.com/show_bug.cgi?id=1387793, but I'm seeing
different address ranges, like [30f00, 31f00), [31000, 32000) and [31800,
32800). There are some more variants, but this morning, they all have one of
the 3 starting points indicated here.
Has there been any development related to this issue?
Should I file another Bugzilla report - seeing as the original one is closed
with INSUFFICIENT_DATA and belongs to a different release?
Is there anything I can do to get around the problem, or at least suppress
the error messages? (They kind of get in my way right now as I'm
experimenting with a framebuffer application.)

They are around the memory usage of the GPU with CMA, they're mostly
informational/informative and not a bug per say, so they're not really
going to be fixed. Updating/opening a new bug will basically just be
ignored/closed because I'm not going to modify/patch the kernel from
upstream based on something that's mildly annoying in the console.

Peter
_______________________________________________
arm mailing list -- ***@lists.fedoraproject.org
To unsubscribe send a