Bug 10840

Summary: netsc520 issues
Product: Drivers Reporter: Roland Kletzing (devzero)
Component: Flash/Memory Technology DevicesAssignee: David Woodhouse (dwmw2)
Status: CLOSED WILL_NOT_FIX    
Severity: normal CC: alan
Priority: P1    
Hardware: All   
OS: Linux   
Kernel Version: 2.6.37-rc7 Subsystem:
Regression: No Bisected commit-id:

Description Roland Kletzing 2008-05-31 01:58:55 UTC
Latest working kernel version: n/a
Earliest failing kernel version:  have seen it with 2.6.25 but maybe older problem
Distribution: opensuse 10.1
Hardware Environment: x86 / P4 / Gibabyte/i915 MoBo
Software Environment:
Problem Description:

Netsc520 has some issue(s)

see 
http://marc.info/?l=linux-kernel&m=120809099119181&w=2
http://marc.info/?l=linux-kernel&m=121219100309366&w=2
http://marc.info/?l=linux-kernel&m=120664963508105&w=2



i repost some oopses here :


reported by Alexey Dobriyan <adobriyan@ gmail.com> :

NetSc520 flash device: 0x100000 at 0x200000
------------[ cut here ]------------
WARNING: at arch/x86/mm/ioremap.c:159 __ioremap_caller+0x15e/0x2c0()
Modules linked in: netsc520(+) [many]
Pid: 13101, comm: modprobe Not tainted 2.6.26-rc4-fat #6

Call Trace:
 [<ffffffff80238042>] warn_on_slowpath+0x58/0x88
 [<ffffffff80238c7e>] ? printk+0x67/0x69
 [<ffffffff80222adc>] __ioremap_caller+0x15e/0x2c0
 [<ffffffffa1fab033>] ? :netsc520:init_netsc520+0x33/0xf4
 [<ffffffff80222c61>] ioremap_nocache+0x12/0x14
 [<ffffffffa1fab033>] :netsc520:init_netsc520+0x33/0xf4
 [<ffffffff80261e35>] sys_init_module+0x9e/0x1b2
 [<ffffffff8020c15a>] system_call_after_swapgs+0x8a/0x8f

---[ end trace dd7999c17a89437b ]---
CFI: Found no netsc520 Flash Bank device at location zero
Creating 4 MTD partitions on "netsc520 Flash Bank":
0x00000000-0x000c0000 : "NetSc520 boot kernel"
mtd: Giving out device 2 to NetSc520 boot kernel
ftl_cs: FTL header not found.
0x000c0000-0x00100000 : "NetSc520 Low BIOS"
mtd: Giving out device 3 to NetSc520 Low BIOS
ftl_cs: FTL header not found.
0x00100000-0x00f80000 : "NetSc520 file system"
mtd: partition "NetSc520 file system" is out of reach -- disabled
mtd: Giving out device 4 to NetSc520 file system
ftl_cs: FTL header not found.
0x00f80000-0x01000000 : "NetSc520 High BIOS"
mtd: partition "NetSc520 High BIOS" is out of reach -- disabled
mtd: Giving out device 5 to NetSc520 High BIOS
ftl_cs: FTL header not found.


reported by me:

> > this one looks more serious (happened with modprobe -r ac),
> > because kernel has problems afterwards (cannot load/anload any other module
> \
> > afterwards) 
> 
> any idea when this started happening?
> Any chance you can snag the output from acpidump
> and attach it to a sighting here?:
> http://bugzilla.kernel.org/enter_bug.cgi?product=ACPI
> 
> thanks,
> -Len

>i searched into this and found that it only happens after modprobe netsc520
>>;modprobe -r netsc520. so maybe this is no acpi problem at all but corruption
>>(sort of?) > introduced by netsc520 module !?
>
>regards
>roland
> 
> > ACPI: ACPI0007:00 is registered as cooling_device0
> > ACPI: Processor [CPU0] (supports 8 throttling states)
> > BUG: unable to handle kernel paging request at 1b563b0c
> > IP: [<c020000f>] acpi_ps_execute_method+0xc4/0x1f9
> > *pde = 00000000
> > Oops: 0002 [#1] SMP
> > Modules linked in: ac(+) sg ahci libata dock sha1_generic michael_mic xcbc
> crc32c \
> > deflate zlib_deflate tgr192 wp512 sha512 seed camellia fcrypt pcbc khazad
> tea \
> > arc4 cast6 cast5 xts gf128mul serpent blowfish sha256_generic md4 cbc
> des_generic \
> > ecb crypto_blkcipher snd_seq_dummy snd_seq snd_timer snd_seq_device snd
> soundcore \
> > parkbd eni suni zatm uPD98402 atm lm70 hwmon ip2 ip2main parport_pc parport
> \
> > 8250_exar_st16c554 8250_accent 8250_hub6 8250_boca 8250_fourport
> ide_generic \
> > hpt366 alim15x3 jmicron pdc202xx_new cs5535 aec62xx atiixp cs5520 it821x
> cs5530 \
> > hpt34x opti621 trm290 amd74xx ns87415 rz1000 pdc202xx_old via82cxxx triflex
> \
> > cmd64x it8213 cy82c693 slc90e66 sc1200 serverworks sis5513 siimage
> ide_cd_mod \
> > cdrom mdacon capifs sctp libcrc32c 9pnet_fd 9pnet sunrpc decnet dccp_ipv6 \
> > dccp_ipv4 dccp ipv6 fuse ide_pci_generic pcnet32 piix mii intel_agp
> ide_core \
> > agpgart ext3 mbcache jbd mptspi mptscsih mptbase scsi_transport_spi sd_mod
> \
> > scsi_mod [last unloaded: sbshc] 
> > Pid: 14520, comm: modprobe Tainted: P         (2.6.25-rc6-git7-default #1)
> > EIP: 0060:[<c020000f>] EFLAGS: 00010206 CPU: 0
> > EIP is at acpi_ps_execute_method+0xc4/0x1f9
> > EAX: 1b563b0c EBX: 00000000 ECX: 0000ffff EDX: 00000093
> > ESI: d41a6240 EDI: c032c726 EBP: 00000000 ESP: df881d98
> > DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> > Process modprobe (pid: 14520, ti=df880000 task=ddb458f0 task.ti=df880000)
> > Stack: ffffffff c032d513 df020c8c df020c84 df021844 d41a6250 d41a6240
> c032c726
> > 00000000 c01fc910 00001001 d41a6240 e085f44c 00000000 c01fc44b e085f44c
> > df020ce8 d41a6a20 d41a6a20 00000004 e085f44c df020ce8 c01ec145 df881dfc
> > Call Trace:
> > [<c01fc910>] acpi_ns_evaluate+0x104/0x1a0
> > [<c01fc44b>] acpi_evaluate_object+0x16f/0x244
> > [<c01ec145>] acpi_evaluate_integer+0x84/0x104
> > [<e085f038>] acpi_ac_get_state+0x22/0x53 [ac]
> > [<e085f20e>] acpi_ac_add+0x63/0x15a [ac]
> > [<c0207f4d>] acpi_device_probe+0x37/0xcd
> > [<c022dd2b>] driver_probe_device+0x9d/0x114
> > [<c022de86>] __driver_attach+0x52/0x87
> > [<c022d36b>] bus_for_each_dev+0x31/0x52
> > [<c022dbde>] driver_attach+0x11/0x13
> > [<c022de34>] __driver_attach+0x0/0x87
> > [<c022da1e>] bus_add_driver+0x91/0x193
> > [<c022e039>] driver_register+0x45/0x9a
> > [<e0863021>] acpi_ac_init+0x21/0x39 [ac]
> > [<c013c1de>] sys_init_module+0x16b6/0x17c9
> > [<c0208280>] acpi_bus_register_driver+0x0/0x38
> > [<c0108a6a>] sys_mmap2+0x62/0x77
> > [<c01059c4>] sysenter_past_esp+0x6d/0xa9
> > =======================
> > Code: c0 b8 01 00 00 00 e8 f9 53 00 00 ff 76 08 8b 46 10 50 83 c0 08 50 68
> 13 d5 \
> > 32 c0 6a ff 00 00 00 c2 2d  c0 68 b0 c2 2d c0 68 fb 00 <00> 00 6a 40 e8 82
> 2a 00 \
> >                 00 83 c4 24 e8 93 fe ff ff 85 c0 89 c5
> > EIP: [<c020000f>] acpi_ps_execute_method+0xc4/0x1f9 SS:ESP 0068:df881d98
> > ---[ end trace f8c13e074fc8ef52 ]---

just tried on 2.6.26-rc4-git2, which rendered my system unusable (need reboot). getting segfaults on ls, vi etc. can post recent dmesg on request. just don`t like to crash my box another time. crash your own box ;) 


Steps to reproduce:
modprobe netsc520;modprobe -r netsc520
Comment 1 Roland Kletzing 2008-10-26 15:03:49 UTC
i see that there was a temporary duplicate at http://bugzilla.kernel.org/show_bug.cgi?id=11226.

anyway - i`m not satisfied with that answer - even if this is a specific driver for an embedded board - it should handle that correctly.

there are thousands of modules in the kernel and very, very few show such erratic behaviour.

suse kerels ship with that module built, btw - so this at least creates hassle on module testing.
Comment 2 Roland Kletzing 2008-10-26 15:18:54 UTC
others complain, too - and spend their time with this: https://bugzilla.novell.com/show_bug.cgi?id=301798
Comment 3 Roland Kletzing 2008-10-26 15:27:56 UTC
one more: http://lkml.org/lkml/2007/7/1/152
Comment 4 Alan 2009-03-26 16:37:24 UTC
This is a driver for an obscure embedded demonstration board. It's also one where there doesn't seem to be a good way to sanity check before probing. Quite why any standard PC distro would ship it is beyond me
Comment 5 Roland Kletzing 2010-12-29 22:02:29 UTC
obscure or not - this problem still exists for 2.6.37-rc7 and this is one of the very very few drivers, which cause trouble on load. i have run another module testing session and again came across this one. (can`t count,how often, in the meantime....)

and furthermore - distros ship it!
ubuntu 10.10 has it and i assume some other distros ship it, too. have at least seen on suse.

furthermore, we have another 2 bug reports from people who apparently did statically compile this into their kernel:
https://bugzilla.kernel.org/show_bug.cgi?id=11226
https://bugzilla.kernel.org/show_bug.cgi?id=12357

and thomas gleixner said: "Actually this driver is complete crap."

so, i would recommend that this driver will either be fixed or marked as broken (some better idea?), so it won`t get unneccessarily built and shipped with every sort of distro. 

whoever is in need of that driver will be able to build it, anyway. but i suspect someone is using 2.6.37 kernel on that old 133Mhz processor. (which got recently EOL`ed, btw: http://support.eurotech-inc.com/developers/documents/PCN10002.pdf )