[Risolto] Kernel panic: bad RIP value

[Ansa89]

Da qualche giorno ottengo dei kernel panic casuali, quando compilo programmi corposi (kernel, mame, mess).
Uso slackware64 13.37 e ho fatto delle prove con kernel 3.1.5, 3.1.10 e 3.2.1.

Purtroppo non riesco a prendere l'intero stack trace, ma riesco a leggere "kernel: Code: Bad RIP value".
C'è qualcosa che posso fare per capire da cosa sono causati questi kernel panic?

[erio]

ansa89,fai un controllo sulla ram installata,da un tread risolto su gentoo forum.

[Ansa89]

Memtest non ha riportato errori .

[erio]

con il 3.2.1 ho avuto problemi perche' nel config avevo abilitato embedded e la voce config_expert che si abilita di conseguenza, problemi con modprobe e vmalloc,tolta la spunta,ora gira tutto bene,magari vedi se hai questa voce abilitata. ma se lo compili e hai errori non credo dipenda dal kernel.

[414N]

Qui un utente dice di non riuscire addirittura ad avviare il sistema per via di questo errore, ed un altro utente gli risponde dicendo che dovrebbe essere un problema del BIOS evitabile passando come argomento

Codice: Seleziona tutto

idle=poll

al kernel all'avvio. Nonostante i post siano del 2004 io farei una prova.

[Ansa89]

L'avevo visto anche io, ma proprio perché era del 2004 avevo rinunciato.
Farò una prova.

[Ansa89]

Ho avviato una compilazione del kernel con un config tutto "y" (in pratica il worst case ever).
Adesso vediamo come si comporta.

Nella documentazione del kernel ho trovato la spiegazione di "idle=poll" e non sono proprio felice di averlo abilitato:

Codice: Seleziona tutto

        idle=           [X86]
                        Format: idle=poll, idle=mwait, idle=halt, idle=nomwait
                        Poll forces a polling idle loop that can slightly
                        improve the performance of waking up a idle CPU, but
                        will use a lot of power and make the system run hot.
                        Not recommended.
                        idle=mwait: On systems which support MONITOR/MWAIT but
                        the kernel chose to not use it because it doesn't save
                        as much power as a normal idle loop, use the
                        MONITOR/MWAIT idle loop anyways. Performance should be
                        the same as idle=poll.
                        idle=halt: Halt is forced to be used for CPU idle.
                        In such case C2/C3 won't be used again.
                        idle=nomwait: Disable mwait for CPU C-states

In pratica avrò un surriscaldamento del processore (che non è amico delle performance) .

[Ansa89]

La compilazione è andata bene fino verso la fine, ma poi:

Codice: Seleziona tutto

  LD      vmlinux.o
  MODPOST vmlinux.o
WARNING: vmlinux.o(.init.text+0x92536): Section mismatch in reference from the function b44_init() to the function .exit.text:b44_pci_exit()
The function __init b44_init() references
a function __exit b44_pci_exit().
This is often seen when error handling in the init function
uses functionality in the exit path.
The fix is often to remove the __exit annotation of
b44_pci_exit() so it may be used outside an exit section.

WARNING: vmlinux.o(.devinit.text+0x1ad4c2): Section mismatch in reference from the function cnb20le_res() to the function .init.text:acpi_os_get_root_pointer()
The function __devinit cnb20le_res() references
a function __init acpi_os_get_root_pointer().
If acpi_os_get_root_pointer is only used by cnb20le_res then
annotate acpi_os_get_root_pointer with a matching annotation.

WARNING: vmlinux.o(.data+0x38f270): Section mismatch in reference from the variable cs5535_mfd_drv to the function .devinit.text:cs5535_mfd_probe()
The variable cs5535_mfd_drv references
the function __devinit cs5535_mfd_probe()
If the reference is valid then annotate the
variable with __init* or __refdata (see linux/init.h) or name the variable:
*driver, *_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

WARNING: vmlinux.o(.data+0x38f278): Section mismatch in reference from the variable cs5535_mfd_drv to the function .devexit.text:cs5535_mfd_remove()
The variable cs5535_mfd_drv references
the function __devexit cs5535_mfd_remove()
If the reference is valid then annotate the
variable with __exit* (see linux/init.h) or name the variable:
*driver, *_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

WARNING: vmlinux.o(.data+0x3f7690): Section mismatch in reference from the variable pch_spi_pcidev to the function .devinit.text:pch_spi_probe()
The variable pch_spi_pcidev references
the function __devinit pch_spi_probe()
If the reference is valid then annotate the
variable with __init* or __refdata (see linux/init.h) or name the variable:
*driver, *_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

WARNING: vmlinux.o(.data+0x3f7698): Section mismatch in reference from the variable pch_spi_pcidev to the function .devexit.text:pch_spi_remove()
The variable pch_spi_pcidev references
the function __devexit pch_spi_remove()
If the reference is valid then annotate the
variable with __exit* (see linux/init.h) or name the variable:
*driver, *_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console

  GEN     .version
  CHK     include/generated/compile.h
  UPD     include/generated/compile.h
  CC      init/version.o
  LD      init/built-in.o
  LD      .tmp_vmlinux1
  KSYM    .tmp_kallsyms1.S
  AS      .tmp_kallsyms1.o
  LD      .tmp_vmlinux2
  KSYM    .tmp_kallsyms2.S
  AS      .tmp_kallsyms2.o
  CHK     include/linux/version.h
  LD      vmlinux
  SYSMAP  System.map
  SYSMAP  .tmp_System.map
  TEST    posttest
  Building modules, stage 2.
  MODPOST 134 modules
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd1c9:      c5 f9 73 de 04          vpsrldq $0x4,%xmm6,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd1eb:      c5 b1 73 fd 0c          vpslldq $0xc,%xmm5,%xmm9
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd1f0:      c5 f9 72 f5 01          vpslld $0x1,%xmm5,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd212:      c5 d1 72 d5 1f          vpsrld $0x1f,%xmm5,%xmm5
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd21b:      c4 c1 51 72 f1 02       vpslld $0x2,%xmm9,%xmm5
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd221:      c4 c1 31 72 d1 1e       vpsrld $0x1e,%xmm9,%xmm9
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd26e:      c5 f9 73 dd 04          vpsrldq $0x4,%xmm5,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd291:      c5 b1 73 fc 0c          vpslldq $0xc,%xmm4,%xmm9
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd296:      c5 f9 72 f4 01          vpslld $0x1,%xmm4,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd2b8:      c5 d9 72 d4 1f          vpsrld $0x1f,%xmm4,%xmm4
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd2c1:      c4 c1 59 72 f1 02       vpslld $0x2,%xmm9,%xmm4
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd2c7:      c4 c1 31 72 d1 1e       vpsrld $0x1e,%xmm9,%xmm9
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd316:      c5 f9 73 dc 04          vpsrldq $0x4,%xmm4,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd338:      c5 b1 73 fb 0c          vpslldq $0xc,%xmm3,%xmm9
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd33d:      c5 f9 72 f3 01          vpslld $0x1,%xmm3,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd35f:      c5 e1 72 d3 1f          vpsrld $0x1f,%xmm3,%xmm3
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd368:      c4 c1 61 72 f1 02       vpslld $0x2,%xmm9,%xmm3
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd36e:      c4 c1 31 72 d1 1e       vpsrld $0x1e,%xmm9,%xmm9
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd3bd:      c5 f9 73 db 04          vpsrldq $0x4,%xmm3,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd3df:      c5 b1 73 fa 0c          vpslldq $0xc,%xmm2,%xmm9
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd3e4:      c5 f9 72 f2 01          vpslld $0x1,%xmm2,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd406:      c5 e9 72 d2 1f          vpsrld $0x1f,%xmm2,%xmm2
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd40f:      c4 c1 69 72 f1 02       vpslld $0x2,%xmm9,%xmm2
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd415:      c4 c1 31 72 d1 1e       vpsrld $0x1e,%xmm9,%xmm9
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd49e:      c5 f9 72 f1 02          vpslld $0x2,%xmm1,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd4a3:      c5 f1 72 d1 1e          vpsrld $0x1e,%xmm1,%xmm1
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd51f:      c4 c1 79 72 f0 02       vpslld $0x2,%xmm8,%xmm0
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd525:      c4 c1 39 72 d0 1e       vpsrld $0x1e,%xmm8,%xmm8
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd5a2:      c5 f9 72 f7 02          vpslld $0x2,%xmm7,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd5a7:      c5 c1 72 d7 1e          vpsrld $0x1e,%xmm7,%xmm7
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd622:      c5 f9 72 f6 02          vpslld $0x2,%xmm6,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd627:      c5 c9 72 d6 1e          vpsrld $0x1e,%xmm6,%xmm6
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd6a1:      c5 f9 72 f5 02          vpslld $0x2,%xmm5,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd6a6:      c5 d1 72 d5 1e          vpsrld $0x1e,%xmm5,%xmm5
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd721:      c5 f9 72 f4 02          vpslld $0x2,%xmm4,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd726:      c5 d9 72 d4 1e          vpsrld $0x1e,%xmm4,%xmm4
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd7b3:      c5 f9 72 f3 02          vpslld $0x2,%xmm3,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd7b8:      c5 e1 72 d3 1e          vpsrld $0x1e,%xmm3,%xmm3
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd84b:      c5 f9 72 f2 02          vpslld $0x2,%xmm2,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd850:      c5 e9 72 d2 1e          vpsrld $0x1e,%xmm2,%xmm2
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd8e3:      c5 f9 72 f1 02          vpslld $0x2,%xmm1,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd8e8:      c5 f1 72 d1 1e          vpsrld $0x1e,%xmm1,%xmm1
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd97a:      c4 c1 79 72 f0 02       vpslld $0x2,%xmm8,%xmm0
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cd980:      c4 c1 39 72 d0 1e       vpsrld $0x1e,%xmm8,%xmm8
Warning: objdump says 6 bytes, but insn_get_length() says 4
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cda15:      c5 f9 72 f7 02          vpslld $0x2,%xmm7,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cda1a:      c5 c1 72 d7 1e          vpsrld $0x1e,%xmm7,%xmm7
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cdaa6:      c5 f9 72 f6 02          vpslld $0x2,%xmm6,%xmm0
Warning: objdump says 5 bytes, but insn_get_length() says 3
Warning: arch/x86/tools/test_get_len found difference at <sha1_transform_avx>:ffffffff810cd120

Warning: ffffffff810cdaab:      c5 c9 72 d6 1e          vpsrld $0x1e,%xmm6,%xmm6
Warning: objdump says 5 bytes, but insn_get_length() says 3

La buona notizia è che non è andato in kernel panic.
Adesso farò una seconda prova .

[Ansa89]

Ok, l'errore precedente sembra essere dovuto ad una cattiva configurazione del kernel; ricompilando il kernel con un cofnig più ridotto non ci sono stati problemi.
Adesso mi viene un dubbio: i kernel panic possono essere causati dal fatto che non ho compilato il supporto per IOMMU?
In particolare:

Memory protection from malicious or misbehaving devices: a device cannot read or write to memory that hasn't been explicitly allocated (mapped) for it. The memory protection is based on the fact that OS running on the CPU (see figure) exclusively controls both the MMU and the IOMMU. The devices are physically unable to circumvent or corrupt configured memory management tables.

[erio]

http://www.phoronix.com/scan.php?page=n ... px=MTAzNzk

[Ansa89]

If you upgraded today to the just-released Linux 3.2 kernel and your Intel system is now having problems booting this new kernel release, you're not alone, but here's a possible workaround.

1) Kernel 3.2.1 (anche se dal changelog non sembra siano stati fatti cambiamenti a IOMMU).
2) Sistema AMD.

[erio]

AMD has published a specification for IOMMU technology in the HyperTransport architecture.[1] Intel has published a specification for IOMMU technology as Virtualization Technology for Directed I/O, abbreviated VT-d.

[Ansa89]

Sì, lo so: il link l'ho postato io e ho chiesto se abilitando questa funzione si può evitare il kernel panic.

[414N]

Nella documentazione del kernel ho trovato la spiegazione di "idle=poll" e non sono proprio felice di averlo abilitato:

Codice: Seleziona tutto

        idle=           [X86]
                        Format: idle=poll, idle=mwait, idle=halt, idle=nomwait
                        Poll forces a polling idle loop that can slightly
                        improve the performance of waking up a idle CPU, but
                        will use a lot of power and make the system run hot.
                        Not recommended.

In pratica avrò un surriscaldamento del processore (che non è amico delle performance) .

Più che surriscaldamento (quello lo avresti con la CPU fissa al 100% sempre) consumerai un po' più di energia perché non metterai a riposo il processore nei periodi idle ma lo manterrai in pausa attiva. Per le performance non credo noterai peggioramenti, anzi: essendo in pausa attiva il processore si sveglia dallo stato idle più velocemente (come scritto sopra).
Riguardo IOMMU non so cosa dirti: io lo metto sempre il supporto nel kernel, anche se non credo di disporre di hardware che lo usi (non trovo nessuna stringa che contenga iommu in /proc o /sys, a parte qualche voce relativa al modulo fglrx).

[slux]

Hai controllato le temperature della CPU e della motherboard?
Se il problema lo hai sempre sotto lunghi stress,magari è quello.