hi, Eli, et al.

It's all in all a very interesting thread :)

I have 2 points of contribution:

1) I hope I didn't miss a sentence in one of the emails "I've searched the 
google for 48 hours for similar problem..."
Are there any other people like yourself in the world ?
(I don't mean you're not unique and interesting guy, but I mean if there are 
any reports on problem like yours... in the kernel mailing lists ? or fedora ? )

2) Also, really to the point, I do not dispute it may be the kernel code to 
blame. But let's localize it:
The problem may be not in the reading the tty/pts, but with input-output 
synchronization.
I have noticed the opening of the device is done with these options:
O_RDWR | O_NOCTTY
I can imagine a situation when your tty does open quickly, but the call also 
may ask the display to be redrawn. 
if there's some text in the output buffer, from previous application, for 
example.
And until writing to console has completed the redraw the open() call is stuck.
And yes,  I know what O_NOCTTY means.

so... on that note, it is possible the problem is with the framebuffer code.

To see if this is true, disable all the framebuffer stuff.
See how it performs.
And get back to us :)

Shabat Shalom.



On Oct 29, 2010, at 11:28, Eli Billauer wrote:

> Just to have SELinux out of the way: I just compiled the kernel with 
> SELinux opted out, and the same freeze behavior remains.
> 
> 
>   Eli
> 
> 
> Eli Billauer wrote:
> 
>> And as a side remark, I've had two null pointer oopses when logging out 
>> the GUI console. The caller seems to be 
>> selinux_socket_unix_stream_connect(). Oopses. I think I'm going to post 
>> a kernel patch which paints the screen background blue.
>> 
>> 
>> 
> 
> -- 
> Web: http://www.billauer.co.il
> 
> _______________________________________________
> Haifux mailing list
> Haifux@haifux.org
> http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux

Attachment: smime.p7s
Description: S/MIME cryptographic signature

_______________________________________________
Haifux mailing list
Haifux@haifux.org
http://hamakor.org.il/cgi-bin/mailman/listinfo/haifux

Reply via email to