[parisc-linux] EISA 100/10 add-on card causes kernel BUG

[parisc-linux] EISA 100/10 add-on card causes kernel BUG

[Rafael E. Herrera]

A while ago I asked about support for the EISA 100/10 network cards. By 
enabling the HP100 driver the card gets detected as eth0.

During boot, at the time when the network interface is set up a kernel 
BUG is generated. Similar behavior with kernel version 2.4.18-pa0, too. 
What follows is the console output:

Linux version 2.4.18-pa5 (raffo@inca) (gcc version 3.0.2 20010829 
(prerelease)) #3 Sat Mar 16 00:29:45 EST 2002
FP[0] enabled: Rev 1 Model 9
The 32-bit Kernel has started...
Determining PDC firmware type: Snake.
model 00003160 00000481 00000000 00000000 77b82c78 ffffffff 00000004 
0000000a 0000000a
vers  00000013
model 9000/715
Total Memory: 64 Mb
pagetable_init
On node 0 totalpages: 16384
zone(0): 16384 pages.
zone(1): 0 pages.
zone(2): 0 pages.
Kernel command line: HOME=/ TERM=linux root=/dev/sda3 
eisa_irq_edge=3,4,5,7,9,10,11,14,15  console=ttyS0 palo_kernel=0/vmlinux
Console: colour dummy device 160x64
Calibrating delay loop... 74.75 BogoMIPS
Memory: 61500k available
Dentry-cache hash table entries: 8192 (order: 4, 65536 bytes)
Inode-cache hash table entries: 4096 (order: 3, 32768 bytes)
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes)
Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes)
Page-cache hash table entries: 16384 (order: 4, 65536 bytes)
Searching for devices...
Found devices:
1. Stinger Optional Graphics (10) at 0xf4000000 [0], versions 0x6, 0x0, 0x77
2. Scorpio Sr. Core BA (11) at 0xf082f000 [2], versions 0x19, 0x0, 0x70
3. Scorpio Sr. Core SCSI (10) at 0xf0825000 [2/0/1], versions 0x19, 0x0, 
0x71
4. Scorpio Sr. Core LAN (802.3) (10) at 0xf0826000 [2/0/2], versions 
0x19, 0x0, 0x72
5. Scorpio Sr. Core HIL (10) at 0xf0821000 [2/0/3], versions 0x19, 0x0, 0x73
6. Scorpio Sr. Core RS-232 (10) at 0xf0823000 [2/0/4], versions 0x19, 
0x0, 0x75
7. Scorpio Sr. Core RS-232 (10) at 0xf0822000 [2/0/5], versions 0x19, 
0x0, 0x75
8. Scorpio Sr. Core Centronics (10) at 0xf0824000 [2/0/6], versions 
0x19, 0x0, 0x74
9. Scorpio Sr. Audio (10) at 0xf1000000 [2/0/8], versions 0x19, 0x0, 0x7b
10. Scorpio Sr. EISA BA (11) at 0xfc000000 [4], versions 0x19, 0x0, 0x76
11. Scorpio Sr.(715/75) (0) at 0xfffbe000 [8], versions 0x316, 0x0, 0x4
12. Memory (1) at 0xfffbf000 [9], versions 0x27, 0x0, 0x9
CPU(s): 1 x PA7100 (PCX-T) at 75.000000 MHz
Asp version 1 at 0xf082f000 found.
LED (ASP-style) display at f0800020 registered
Mongoose EISA Adapter found at 0xfc000000
EISA EEPROM at 0xf0810400
Enumerating EISA bus
EISA slot 1: HWP1990 EISA IRQ 3 <4>IRQ 3 polarity configured twice (last 
to level)

POSIX conformance testing by UNIFIX
Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
Starting kswapd
Journalled Block Device driver loaded
parport_init_chip: enhanced parport-modes not supported.
parport0: PC-style at 0xf0824800, irq 88 [PCSPP]
STI byte mode ROM at f4000000, hpa=f4000000
STI byte mode ROM, id 27f12392-40a00499, conforms to spec rev. 8.02
STI device: HPA1991AC19
Console: switching to colour frame buffer device 160x64
fb0: stifb 1280x1024-8 frame buffer device, id: 27f12392, mmio: 0xf4100000
pty: 256 Unix98 ptys configured
Serial driver version 5.05c (2001-07-08) with MANY_PORTS SHARE_IRQ 
SERIAL_PCI enabled
ttyS00 at iomem 0xf0823800 (irq = 90) is a 16550A
ttyS01 at iomem 0xf0822800 (irq = 89) is a 16550A
Found HIL at 0xf0821000, IRQ 94
HIL: no keyboard present.
lp0: using parport0 (interrupt-driven).
Generic RTC Driver v1.02 05/27/1999 Sam Creasey (sammy@oh.verio.com)
block: 128 slots per queue, batch=32
RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 50MHz system bus speed for PIO modes; override with idebus=xx
hp100: eth0: Using (slow) programmed i/o mode.
hp100: eth0: HP J2577 at 0x1c38, IRQ 3, EISA bus, 128k SRAM (rx/tx 75%).
hp100: eth0: Adapter is attached to 10Mb/s network.
loop: loaded (max 8 devices)
Found i82596 at 0xf0826000, IRQ 87
82596.c: MAC of HP700 LAN read from EEPROM
eth1: Couldn't get consistent shared memory
eth1: 82596 at 0xf0826000, 08 00 09 87 89 28 IRQ 87.
82596.c $Revision: 1.29 $
SCSI subsystem driver Revision: 1.00
53c700: consistent memory allocation failed
53c700: Version 2.7 By James.Bottomley@HansenPartnership.com
scsi0: 53c700 rev 0
scsi0 : LASI SCSI 53c700
   Vendor: DEC       Model: RZ26L    (C) DEC  Rev: 440C
   Type:   Direct-Access                      ANSI SCSI revision: 02
Attached scsi disk sda at scsi0, channel 0, id 6, lun 0
scsi0: (6:0) Enabling Tag Command Queuing
SCSI device sda: 2050860 512-byte hdwr sectors (1050 MB)
Partition check:
  sda: sda1 sda2 sda3
Lasi Harmony Audio driver V0.9a, h/w id 20, rev. 0 at 0xf1000000, IRQ 82
sticonsole_init: searching for STI ROMs
md: linear personality registered as nr 1
md: raid0 personality registered as nr 2
md: raid1 personality registered as nr 3
md: raid5 personality registered as nr 4
raid5: measuring checksumming speed
    8regs     : 12246.800 MB/sec
    8regs_prefetch:    55.600 MB/sec
    32regs    :    56.800 MB/sec
    32regs_prefetch:    56.800 MB/sec
raid5: using function: 8regs (12246.800 MB/sec)
md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
md: Autodetecting RAID arrays.
md: autorun ...
md: ... autorun DONE.
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
IP: routing cache hash table of 512 buckets, 4Kbytes
TCP: Hash tables configured (established 4096 bind 8192)
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
VFS: Mounted root (ext2 filesystem) readonly.
NOT FREEING INITMEM (386k)
INIT: version 2.84 booting
Activating swap.
Adding Swap: 124800k swap-space (priority -1)
Checking root file system...
fsck 1.26 (3-Feb-2002)
/dev/sda3: clean, 37561/110656 files, 129875/220926 blocks
System time was Sat Mar 16 05:42:22 UTC 2002.
Setting the System Clock using the Hardware Clock as reference...
System Clock set. System local time is now Sat Mar 16 05:42:35 UTC 2002.
Checking all file systems...
fsck 1.26 (3-Feb-2002)
Setting kernel variables.
Loading the saved-state of the serial devices...
Cannot set serial info: Address already in use
/dev/ttyS0 at 0x0000 (irq = 90) is a 16550A
Cannot set serial info: Address already in use
/dev/ttyS1 at 0x0000 (irq = 89) is a 16550A
Mounting local filesystems...
nothing was mounted
Cleaning: /etc/network/ifstate.
Setting up IP spoofing protection: rp_filter.
Configuring network interfaces:

kernel BUG at printk.c:489!
Linux version 2.4.18-pa5 (raffo@inca) (gcc version 3.0.2 20010829 
(prerelease)) #3 Sat Mar 16 00:29:45 EST 2002
FP[0] enabled: Rev 1 Model 9
The 32-bit Kernel has started...
Determining PDC firmware type: Snake.
model 00003160 00000481 00000000 00000000 77b82c78 ffffffff 00000004 
0000000a 0000000a
vers  00000013
model 9000/715
Total Memory: 64 Mb
pagetable_init
On node 0 totalpages: 16384
zone(0): 16384 pages.
zone(1): 0 pages.
zone(2): 0 pages.
Kernel command line: HOME=/ TERM=linux root=/dev/sda3 
eisa_irq_edge=3,4,5,7,9,10,11,14,15  console=ttyS0 palo_kernel=0/vmlinux
Console: colour dummy device 160x64
Calibrating delay loop... 74.75 BogoMIPS
Memory: 61500k available
Dentry-cache hash table entries: 8192 (order: 4, 65536 bytes)
Inode-cache hash table entries: 4096 (order: 3, 32768 bytes)
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes)
Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes)
Page-cache hash table entries: 16384 (order: 4, 65536 bytes)
Searching for devices...
Found devices:
1. Stinger Optional Graphics (10) at 0xf4000000 [0], versions 0x6, 0x0, 0x77
2. Scorpio Sr. Core BA (11) at 0xf082f000 [2], versions 0x19, 0x0, 0x70
3. Scorpio Sr. Core SCSI (10) at 0xf0825000 [2/0/1], versions 0x19, 0x0, 
0x71
4. Scorpio Sr. Core LAN (802.3) (10) at 0xf0826000 [2/0/2], versions 
0x19, 0x0, 0x72
5. Scorpio Sr. Core HIL (10) at 0xf0821000 [2/0/3], versions 0x19, 0x0, 0x73
6. Scorpio Sr. Core RS-232 (10) at 0xf0823000 [2/0/4], versions 0x19, 
0x0, 0x75
7. Scorpio Sr. Core RS-232 (10) at 0xf0822000 [2/0/5], versions 0x19, 
0x0, 0x75
8. Scorpio Sr. Core Centronics (10) at 0xf0824000 [2/0/6], versions 
0x19, 0x0, 0x74
9. Scorpio Sr. Audio (10) at 0xf1000000 [2/0/8], versions 0x19, 0x0, 0x7b
10. Scorpio Sr. EISA BA (11) at 0xfc000000 [4], versions 0x19, 0x0, 0x76
11. Scorpio Sr.(715/75) (0) at 0xfffbe000 [8], versions 0x316, 0x0, 0x4
12. Memory (1) at 0xfffbf000 [9], versions 0x27, 0x0, 0x9
CPU(s): 1 x PA7100 (PCX-T) at 75.000000 MHz
Asp version 1 at 0xf082f000 found.
LED (ASP-style) display at f0800020 registered
Mongoose EISA Adapter found at 0xfc000000
EISA EEPROM at 0xf0810400
Enumerating EISA bus
EISA slot 1: HWP1990 EISA IRQ 3 <4>IRQ 3 polarity configured twice (last 
to level)

POSIX conformance testing by UNIFIX
Linux NET4.0 for Linux 2.4
Based upon Swansea University Computer Society NET3.039
Initializing RT netlink socket
Starting kswapd
Journalled Block Device driver loaded
parport_init_chip: enhanced parport-modes not supported.
parport0: PC-style at 0xf0824800, irq 88 [PCSPP]
STI byte mode ROM at f4000000, hpa=f4000000
STI byte mode ROM, id 27f12392-40a00499, conforms to spec rev. 8.02
STI device: HPA1991AC19
Console: switching to colour frame buffer device 160x64
fb0: stifb 1280x1024-8 frame buffer device, id: 27f12392, mmio: 0xf4100000
pty: 256 Unix98 ptys configured
Serial driver version 5.05c (2001-07-08) with MANY_PORTS SHARE_IRQ 
SERIAL_PCI enabled
ttyS00 at iomem 0xf0823800 (irq = 90) is a 16550A
ttyS01 at iomem 0xf0822800 (irq = 89) is a 16550A
Found HIL at 0xf0821000, IRQ 94
HIL: no keyboard present.
lp0: using parport0 (interrupt-driven).
Generic RTC Driver v1.02 05/27/1999 Sam Creasey (sammy@oh.verio.com)
block: 128 slots per queue, batch=32
RAMDISK driver initialized: 16 RAM disks of 4096K size 1024 blocksize
Uniform Multi-Platform E-IDE driver Revision: 6.31
ide: Assuming 50MHz system bus speed for PIO modes; override with idebus=xx
hp100: eth0: Using (slow) programmed i/o mode.
hp100: eth0: HP J2577 at 0x1c38, IRQ 3, EISA bus, 128k SRAM (rx/tx 75%).
hp100: eth0: Adapter is attached to 10Mb/s network.
loop: loaded (max 8 devices)
Found i82596 at 0xf0826000, IRQ 87
82596.c: MAC of HP700 LAN read from EEPROM
eth1: Couldn't get consistent shared memory
eth1: 82596 at 0xf0826000, 08 00 09 87 89 28 IRQ 87.
82596.c $Revision: 1.29 $
SCSI subsystem driver Revision: 1.00
53c700: consistent memory allocation failed
53c700: Version 2.7 By James.Bottomley@HansenPartnership.com
scsi0: 53c700 rev 0
scsi0 : LASI SCSI 53c700
   Vendor: DEC       Model: RZ26L    (C) DEC  Rev: 440C
   Type:   Direct-Access                      ANSI SCSI revision: 02
Attached scsi disk sda at scsi0, channel 0, id 6, lun 0
scsi0: (6:0) Enabling Tag Command Queuing
SCSI device sda: 2050860 512-byte hdwr sectors (1050 MB)
Partition check:
  sda: sda1 sda2 sda3
Lasi Harmony Audio driver V0.9a, h/w id 20, rev. 0 at 0xf1000000, IRQ 82
sticonsole_init: searching for STI ROMs
md: linear personality registered as nr 1
md: raid0 personality registered as nr 2
md: raid1 personality registered as nr 3
md: raid5 personality registered as nr 4
raid5: measuring checksumming speed
    8regs     : 12246.800 MB/sec
    8regs_prefetch:    55.600 MB/sec
    32regs    :    56.800 MB/sec
    32regs_prefetch:    56.800 MB/sec
raid5: using function: 8regs (12246.800 MB/sec)
md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
md: Autodetecting RAID arrays.
md: autorun ...
md: ... autorun DONE.
NET4: Linux TCP/IP 1.0 for NET4.0
IP Protocols: ICMP, UDP, TCP, IGMP
IP: routing cache hash table of 512 buckets, 4Kbytes
TCP: Hash tables configured (established 4096 bind 8192)
NET4: Unix domain sockets 1.0/SMP for Linux NET4.0.
VFS: Mounted root (ext2 filesystem) readonly.
NOT FREEING INITMEM (386k)
Adding Swap: 124800k swap-space (priority -1)
kernel BUG at printk.c:489!
kernel BUG at printk.c:489!
WARNING! Stack pointer and cr30 do not correspond!
Dumping virtual address stack instead

Dumping Stack from 0x24000000 to 0x24000240:
WARNING! Stack pointer and cr30 do not correspond!
Dumping virtual address stack instead

[a lot of these ...]

Dumping Stack from 0x202f8000 to 0x202f8b40:
WARNING! Stack pointer and cr30 do not correspond!
Dumping virtual address stack instead

Dumping Stack from 0x202f8000 to 0x202f8f80:
WARNING! Stack pointer and cr30 do not correspond!
Dumping virtual address stack instead

Dumping Stack from 0x202f8000 to 0x202f93c0:
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!
WARNING! Stack pointer and cr30 do not correspond!


-- 
      Rafael

Re: [parisc-linux] EISA 100/10 add-on card causes kernel BUG

[Grant Grundler]

"Rafael E. Herrera" wrote:
...
> WARNING! Stack pointer and cr30 do not correspond!

A few times, this error (it's not a warning) meant the 
driver was attempting to access user space data directly.
This is wrong and unfortunately it *usually* (but not always)
works on x86 arch. The right way to access user space data
is via copy_from_user() or write asm code that explicitly
references Space Registers (eg sr3).

grant

Re: [parisc-linux] EISA 100/10 add-on card causes kernel BUG

[Matthew Wilcox]

On Sat, Mar 16, 2002 at 11:42:31AM -0700, Grant Grundler wrote:
> works on x86 arch. The right way to access user space data
> is via copy_from_user() or write asm code that explicitly
> references Space Registers (eg sr3).

Let me just explicitly discourage doing this (or, if you must, document
it in Documentation/parisc/registers).  Use of sr3 to access userspace
is subject to change and it'd be nice not to have to check every driver
for use.  It's bad enough that sba & ccio use sr1, but at least we found
those :-)

I cant imagine a good reason for using sr3.  copy_{to,from}_user are
portable and obvious.

-- 
Revolutions do not require corporate support.

Re: [parisc-linux] EISA 100/10 add-on card causes kernel BUG

[Grant Grundler]

Matthew Wilcox wrote:
> Let me just explicitly discourage doing this (or, if you must, document
> it in Documentation/parisc/registers).  Use of sr3 to access userspace
> is subject to change and it'd be nice not to have to check every driver
> for use. 

Yeah - you are right

> It's bad enough that sba & ccio use sr1, but at least we found
> those :-)

oh - I forgot about those for filling the IO pdir entries...
that need to change in order to support zero copy?

We somehow need to pass the spaceID of the userspace buffer
to dma code if we want DMA to be cache coherent.
I don't have any good ideas for this.

> I cant imagine a good reason for using sr3.  copy_{to,from}_user are
> portable and obvious.

agreed.

thanks,
grant

--9B420482D.1016343457/dsl2.external.hp.com--

Re: [parisc-linux] EISA 100/10 add-on card causes kernel BUG

[Matthew Wilcox]

On Sat, Mar 16, 2002 at 10:38:38PM -0700, Grant Grundler wrote:
> oh - I forgot about those for filling the IO pdir entries...
> that need to change in order to support zero copy?
> 
> We somehow need to pass the spaceID of the userspace buffer
> to dma code if we want DMA to be cache coherent.
> I don't have any good ideas for this.

My understanding is that DMA to userspace is to pages in the page cache
only.  As such, the normal mechanisms for dcache coherency should be used.
That is, DMA to a kernel view of the page and call flush_dcache_page
after touching it.  My proposed scheme for kmap/kunmap implementation
would make flush_dcache_page a no-op, since the kernel would have a view
of the page which is automatically cache coherent with userspace's view.

-- 
Revolutions do not require corporate support.

Re: [parisc-linux] EISA 100/10 add-on card causes kernel BUG

[Grant Grundler]

Matthew Wilcox wrote:
> My understanding is that DMA to userspace is to pages in the page cache
> only.

NICs use the page cache? or just for ZC?
block IO certainly uses the page cache.

> As such, the normal mechanisms for dcache coherency should be used.
> That is, DMA to a kernel view of the page and call flush_dcache_page
> after touching it.

For page cache, ok. Then ccio/sba implicitly use whatever SR the
kernel is supposed to use and we don't need to pass the parameter.
If NIC's use page cache for ZC, that would make life simple.

> My proposed scheme for kmap/kunmap implementation
> would make flush_dcache_page a no-op, since the kernel would have a view
> of the page which is automatically cache coherent with userspace's view.

I'm liking that idea alot.
I've really never been comfortable with the two views not being coherent
since it will expose more kernel bugs that other arches don't see.


thanks,
grant

Re: [parisc-linux] EISA 100/10 add-on card on a 715/80

[Rafael E. Herrera]

Hello,

This time a brought a 715/80 model to test the EISA network card. I do 
not get a kernel BUG.

The network interface is brought up, but I can't connect to the rest of 
the machines in my lan.

I connected the card's 10 BaseT port to my switch.

While trying to ping from the parisc box to my linux PC I get this from 
tcpdump, *.2 is the PC and *.21 is the parisc:

22:04:31.926833 192.168.0.21 > 192.168.0.2: icmp: echo request (DF)
22:04:31.926903 192.168.0.2 > 192.168.0.21: icmp: echo reply [tos 0x10]
22:04:32.926816 192.168.0.21 > 192.168.0.2: icmp: echo request (DF)
22:04:32.926913 192.168.0.2 > 192.168.0.21: icmp: echo reply [tos 0x10]
22:04:33.926793 192.168.0.21 > 192.168.0.2: icmp: echo request (DF)
22:04:33.926885 192.168.0.2 > 192.168.0.21: icmp: echo reply [tos 0x10]
22:04:34.926769 192.168.0.21 > 192.168.0.2: icmp: echo request (DF)
22:04:34.926858 192.168.0.2 > 192.168.0.21: icmp: echo reply [tos 0x10]

Any comments?

P.S. Is there a way to make the hp100 driver bind to eth1 and the built 
in interface bind to eth0?

Thanks.
-- 
      Rafael