Two bugs in fbcon?

Two bugs in fbcon?

[Ondrej Zajicek]

(i resent this email as previous one seems to be lost, so i am sorry
in case you get both ones)

Hello

I found that in  kernel 2.6.16.16 is some very strange behavior in fbcon.
I added some printks to some functions in fbcon and i found this:

I am on vt0, i have active vt0-vt3. When i modprobe fbcon (having fbdev
driver inserted before), functions fbcon.c/fbcon_init() and
fbcon.c/updatescrollmode() are called for each active vt.

Order is:
fbcon_init(vt0)
updatescrollmode(vt0)
updatescrollmode(vt1)
fbcon_init(vt1)
updatescrollmode(vt2)
fbcon_init(vt2)
updatescrollmode(vt3)
fbcon_init(vt3)

for vt1-vt3 updatescrollmode is called before fbcon_init which caused that
associated struct display::vrows value is computed (in updatescrollmode()
- fbcon.c:1970) from old vc->vc_rows (which is updated in fbcon_init()).

Now second strangeness - I am still on vt0 (immediately after modprobe
fbcon), but during calls to bitblit.c/bit_cursor() display structure
associated with vt3 is used (in bitblit.c:244 - ops->p). Combined with
previous strangeness (bad vrows values in vt1-vt3) i get cursor visible
on bad row.

When i switch to another vt or switch resolution, then this is corrected
(next calls to bitblit.c/bit_cursor() are using correct display
structure).

I don't understand fbcon enough to correct this problem (and be sure not to
make some other :-)).


-- 
Elen sila lumenn' omentielvo

Ondrej 'SanTiago' Zajicek (email: santiago@mail.cz, jabber: santiago@njs.netlab.cz)
OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net)
"To err is human -- to blame it on a computer is even more so."


-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Two bugs in fbcon?

[Ondrej Zajicek]

[-- Attachment #1: Type: text/plain, Size: 1562 bytes --]

Hello

I found that in  kernel 2.6.16.16 is some very strange behavior in fbcon.
I added some printks to some functions in fbcon and i found this:

I am on vt0, i have active vt0-vt3. When i modprobe fbcon (having fbdev
driver inserted before), functions fbcon.c/fbcon_init() and
fbcon.c/updatescrollmode() are called for each active vt.

Order is:
fbcon_init(vt0)
updatescrollmode(vt0)
updatescrollmode(vt1)
fbcon_init(vt1)
updatescrollmode(vt2)
fbcon_init(vt2)
updatescrollmode(vt3)
fbcon_init(vt3)

for vt1-vt3 updatescrollmode is called before fbcon_init which caused that
associated struct display::vrows value is computed (in updatescrollmode()
- fbcon.c:1970) from old vc->vc_rows (which is updated in fbcon_init()).

Now second strangeness - I am still on vt0 (immediately after modprobe fbcon),
but during calls to bitblit.c/bit_cursor display structure associated with
vt3 is used (in bitblit.c:244 - ops->p). Combined with previous strangeness
(bad vrows values in vt1-vt3) i get cursor visible on bad row.

When i switch to another vt or switch resolution, then this is corrected 
(next calls to bitblit.c/bit_cursor are using correct display structure.).

I don't understand fbcon enough to correct this problem (and be sure not to
make some other :-)).

-- 
Elen sila lumenn' omentielvo

Ondrej 'SanTiago' Zajicek (email: santiago@mail.cz, jabber: santiago@njs.netlab.cz)
OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net)
"To err is human -- to blame it on a computer is even more so."

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 197 bytes --]

Re: Two bugs in fbcon?

[Antonino A. Daplas]

Ondrej Zajicek wrote:
> Hello
> 
> I found that in  kernel 2.6.16.16 is some very strange behavior in fbcon.
> I added some printks to some functions in fbcon and i found this:
> 
> I am on vt0, i have active vt0-vt3. When i modprobe fbcon (having fbdev
> driver inserted before), functions fbcon.c/fbcon_init() and
> fbcon.c/updatescrollmode() are called for each active vt.
> 
> Order is:
> fbcon_init(vt0)
> updatescrollmode(vt0)
> updatescrollmode(vt1)
> fbcon_init(vt1)
> updatescrollmode(vt2)
> fbcon_init(vt2)
> updatescrollmode(vt3)
> fbcon_init(vt3)
> 
> for vt1-vt3 updatescrollmode is called before fbcon_init which caused that
> associated struct display::vrows value is computed (in updatescrollmode()
> - fbcon.c:1970) from old vc->vc_rows (which is updated in fbcon_init()).
> 
> Now second strangeness - I am still on vt0 (immediately after modprobe fbcon),
> but during calls to bitblit.c/bit_cursor display structure associated with
> vt3 is used (in bitblit.c:244 - ops->p). Combined with previous strangeness
> (bad vrows values in vt1-vt3) i get cursor visible on bad row.
> 
> When i switch to another vt or switch resolution, then this is corrected 
> (next calls to bitblit.c/bit_cursor are using correct display structure.).
> 
> I don't understand fbcon enough to correct this problem (and be sure not to
> make some other :-)).
> 

Might be fixed in Linus's latest git tree.  Try applying this changeset:

http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=1a37d5f51020b9967d4b4094f6d627d0873a6469

Tony



-------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642

Re: Two bugs in fbcon?

[Ondrej Zajicek]

[-- Attachment #1: Type: text/plain, Size: 930 bytes --]

On Sat, May 20, 2006 at 08:30:06AM +0800, Antonino A. Daplas wrote:
> Ondrej Zajicek wrote:
> > Now second strangeness - I am still on vt0 (immediately after modprobe fbcon),
> > but during calls to bitblit.c/bit_cursor display structure associated with
> > vt3 is used (in bitblit.c:244 - ops->p). Combined with previous strangeness
> > (bad vrows values in vt1-vt3) i get cursor visible on bad row.
> 
> Might be fixed in Linus's latest git tree.  Try applying this changeset:
> 
> http://www.kernel.org/git/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=1a37d5f51020b9967d4b4094f6d627d0873a6469
> 
> Tony

Yes, this corrects it. Thank you.

-- 
Elen sila lumenn' omentielvo

Ondrej 'SanTiago' Zajicek (email: santiago@mail.cz, jabber: santiago@njs.netlab.cz)
OpenPGP encrypted e-mails preferred (KeyID 0x11DEADC3, wwwkeys.pgp.net)
"To err is human -- to blame it on a computer is even more so."

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 197 bytes --]