Bug 48279 - [bktr] Brooktre878 may cause freeze
Summary: [bktr] Brooktre878 may cause freeze
Status: Closed Overcome By Events
Alias: None
Product: Base System
Classification: Unclassified
Component: kern (show other bugs)
Version: 4.7-RELEASE
Hardware: Any Any
: Normal Affects Only Me
Assignee: freebsd-bugs (Nobody)
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2003-02-14 12:20 UTC by Peter
Modified: 2022-03-17 01:39 UTC (History)
1 user (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Peter 2003-02-14 12:20:03 UTC
	Take a i386 computer, install a PCI-Brooktre878 card.
	Add 'bktr' device, reboot. Use the computer for a while with
        different applications.
        Then start fxtv-1.03.. FreeZe!

Fix: 

1) Kernel source: Proberbly allocation of memory, or initialisation.
	2) Behaviour:     Run fxtv at once when x11 is started.
Comment 1 Jens Schweikhardt freebsd_committer freebsd_triage 2003-03-29 14:44:25 UTC
State Changed
From-To: open->feedback

This is a known bug in the interaction of brooktree chips with various 
chipset (IS, VIA, UMC, ALI, and some OPTI). It's OS independent. Please 
refer to http://www.hauppauge.com/html/lockups.htm for a list of issues 
and possibilities on how you *may* resolve them.
Comment 2 Jens Schweikhardt 2003-03-29 17:03:51 UTC
Peter,

On Sat, Mar 29, 2003 at 05:01:11PM +0100, Peter B wrote:
# Jens Schweikhardt wrote:
# >Synopsis: Brooktre878 (bktr) may cause freeze
# >
# >State-Changed-From-To: open->feedback
# >State-Changed-By: schweikh
# >State-Changed-When: Sat Mar 29 06:44:25 PST 2003
# >State-Changed-Why: 
# >This is a known bug in the interaction of brooktree chips with various
# >chipset (IS, VIA, UMC, ALI, and some OPTI). It's OS independent. Please
# >refer to http://www.hauppauge.com/html/lockups.htm for a list of issues
# >and possibilities on how you *may* resolve them.
# 
# The reason I wonder wheather it's really a hardware issue.

I had the same problem. I even swapped the bt848 card for a bt878 card.
Then I booted linux on the same system. Same effect: hard freeze
(keyboard LEDs don't change) anywhere from 5min to three days after
launching fxtv. From all I read on the Hauppauge page, it is a HW
incompatibility of the chipset. Which chipset is on your board? If it's
anything other than intel, try swapping the board and see if the problem
persists.

# Is that if I boot
# my computer and start XFree86 (4.2.0) and then fxtv *right away* there is no
# freezes whatsoever, not even after weeks of uptime and frequent use of fxtv in
# all modes. (I'm considering putting ffmpeg in /etc/rc.local)
# 
# However if I boot my computer and start XFree86 (4.2.0) and then start mozilla,
# some xterm's, use ssh.. close some xterm's. And after say 5 minutes start fxtv.
# Instant freeze..
# 
# This made me to conclude it's some kind of initialisation error. And that it
# might have to do with allocation of resources that will not be properly 
# allocated at a later time. It might be kernel memory.

This is not enough information to make a case either way. The symptom
looks very much like a PCI bus lockup, something you can't trigger in
software. I'm not saying it must be HW, but after all my efforts and
investigations, it's by far the most probable.

Regards,

	Jens
-- 
Jens Schweikhardt http://www.schweikhardt.net/
SIGSIG -- signature too long (core dumped)
Comment 3 Jens Schweikhardt 2003-03-30 11:36:56 UTC
On Sat, Mar 29, 2003 at 07:02:55PM +0100, Peter B wrote:
...
# >I had the same problem. I even swapped the bt848 card for a bt878 card.
# >Then I booted linux on the same system. Same effect: hard freeze
# >(keyboard LEDs don't change) anywhere from 5min to three days after
# >launching fxtv. 
# 
# For me it either freezes at once, or work indefinitly. And I can at all times
# avoid the freeze by running fxtv right away and quit the application.
# 
# >From all I read on the Hauppauge page, it is a HW
# >incompatibility of the chipset. Which chipset is on your board? If it's
# >anything other than intel, try swapping the board and see if the problem
# >persists.
# 
# Motherboard:  Asus A7V333-R  (Raid,Firewire-TSB43AB21,Sound-CMI8738)

This board has a VIA KT333 North bridge...

# CPU:          AMD XP1800 (no overclocking)
# AGP:          ATI Technologies RV200 Radeon QW
# Cards:        0 19160 AIC-7892B Ultra160 SCSI Adapter
#               1 Bt878 MediaStream Controller
#               2 RTL8029 NE2000 compatible Ethernet
#               3 RTL8029 NE2000 compatible Ethernet
#               4 <empty>
# 
# >This is not enough information to make a case either way. The symptom
# >looks very much like a PCI bus lockup, something you can't trigger in
# >software. I'm not saying it must be HW, but after all my efforts and
# >investigations, it's by far the most probable.
# 
# If manufactors could straighten out their chips pre-production..F00F..CMD640 :)
# 
# I'll see if a 1 second grab in /etc/rc.local will workaround it in the
# meantime.

It would be very helpful if you could boot linux on your system and try
running kwintv or xawtv. You do not need an extra partition if you use
Knoppix Linux, which is a read-only Linux that runs completely from CD
and RAM. Grab an ISO Image from http://www.knoppix.com/

PS: please CC: copies to gnats with the subject intact (see this mail's
    header) so they can be added to the database.

Regards,

	Jens
-- 
Jens Schweikhardt http://www.schweikhardt.net/
SIGSIG -- signature too long (core dumped)
Comment 4 Peter 2003-06-25 20:44:04 UTC
Regarding PR 48279. I cought an kernel trap. Usually it just freeze, and that's
it. I hope this could be helpfull.

It might be a pci lookup. But the behaviour on my system suggest initialisation
fault. As I can run fxtv for months without any hickups. The only catch is that
the very first thing one have to do is to start xfree86, and then fxtv
immediate after boot.

Another observation I did recently was that running ffmpeg right after boot
freeze the machine. But starting xfree86, and then fxtv. After that ffmpeg runs
without problems.

/var/log/messages:
  Fatal trap 12: page fault while in kernel mode
  fault virtual address   = 0xcddedbe4
  fault code              = supervisor read, page not present
  instruction pointer     = 0x8:0xc02241cb
  stack pointer           = 0x10:0xcddddb30
  frame pointer           = 0x10:0xcddddd60
  code segment            = base rx0, limit 0xfffff, type 0x1b
                          = DPL 0, pres 1, def32 1, gran 1
  processor eflags        = interrupt enabled, resume, IOPL = 0
  current process         = 750 (fxtv)
  interrupt mask          = none
  trap number             = 12
  panic: page fault

  syncing disks... 24 2 1 1 1 1
Comment 5 Mark Linimon freebsd_committer freebsd_triage 2004-08-31 01:07:58 UTC
State Changed
From-To: feedback->suspended

It sounds as though there may be hardware incompatabilities here 
that are non-FreeBSD-related.
Comment 6 Eitan Adler freebsd_committer freebsd_triage 2018-05-20 23:52:20 UTC
For bugs matching the following conditions:
- Status == In Progress
- Assignee == "bugs@FreeBSD.org"
- Last Modified Year <= 2017

Do
- Set Status to "Open"
Comment 7 Ed Maste freebsd_committer freebsd_triage 2022-03-17 01:39:59 UTC
bktr has been removed from FreeBSD