#11734 LOW Not Tri: OFW XO-1.5 - funny display in emacs

Zarro Boogs per Child bugtracker at laptop.org
Mon Feb 11 18:30:10 EST 2013


#11734: OFW XO-1.5 - funny display in emacs
-------------------------------------------+--------------------------------
           Reporter:  wmb at firmworks.com    |       Owner:  wmb at firmworks.com
               Type:  defect               |      Status:  new              
           Priority:  low                  |   Milestone:  Not Triaged      
          Component:  ofw - open firmware  |     Version:  not specified    
         Resolution:                       |    Keywords:                   
        Next_action:  never set            |    Verified:  0                
Deployment_affected:                       |   Blockedby:                   
           Blocking:                       |  
-------------------------------------------+--------------------------------

Comment(by Quozl):

 Removing all alarms does not change the symptom.

 Symptom comes back if the write to the UART transmit holding register is
 disabled:
 {{{
 ok fallback-out-ih iselect
 ok support-ih iselect
 ok patch 2drop uart! write
 }}}

 With the write disabled, symptom goes away if a 1 ms delay per write is
 added in place of ''uart!''.  Therefore the problem is timing critical.

 Capturing the byte sequence emitted by emacs and replaying it from a
 buffer reproduces the symptom.  This excludes emacs.

 Repeating the byte sequence rapidly shows display noise near the top line;
 a white bar about half the height of a text line.

 The display noise can be duplicated with:
 {{{
 \ same colours as emacs
 " "(1b)[1;1H"(1b)[40m"(1b)[J" type many
 \ best colours for viewing
 " "(1b)[1;1H"(1b)[37m"(1b)[40m"(1b)[J" type many
 \ a wider noise band occurs if clearing lower on the screen
 " "(1b)[26;1H"(1b)[37m"(1b)[40m"(1b)[J" type many
 }}}

 Disabling the use of the graphics accelerator for delete lines avoids the
 symptom:
 {{{
 Type 'i' to interrupt stand-init sequence
 Interacting
 ok dev /pci/display at 1  patch noop gp-install set-terminal  dend  resume
 }}}

 But this could be a timing effect.

-- 
Ticket URL: <http://dev.laptop.org/ticket/11734#comment:2>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system


More information about the Bugs mailing list