PCI pci_{un}register_driver interface

From: Petr Vandrovec (VANDROVE@vc.cvut.cz)
Date: Mon Jan 10 2000 - 22:41:26 EST


Hi Martin,
  I just updated my matroxfb to new pci_{un}register_driver interface
and I'd like to ask few questions:
  (1) probe callback should return 0 on failure and !=0 on success.
      Is success value somewhat standardized for future compatibility?
      Why it is not 0=success, !=0 error with returning errno.h compatible
      values (-ENOMEM, -ENXIO, ...)?
  (2) What should suspend & resume callback do? I did not find any
      use in current (2.3.39-pre2) kernel. Should I powerdown/powerup device
      on these callback or what should I do (Matroxes have ACPI power
      states 0 and 3 only)?
  (3a) Can probe & remove callback sleep or should I spawn init/cleanup
      thread when I need sleep (non-atomic kmalloc and so on? (but doing
      cleanup with thread is impossible in remove callback)
  (3b) Should not you protect pci_for_each_dev() and pci_remove_device()
      with some mutex (or add an usecount to pci_dev)? Are there some
      other conditions (except that currently no-one calls
      pci_remove_device()) which make them mutually exclusive (for
      example holding kernel_lock when doing these pci calls...).
  That's all for now.
                                            Thanks,
                                                Petr Vandrovec
                                                vandrove@vc.cvut.cz
                                                      

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Sat Jan 15 2000 - 21:00:16 EST