Bug 9721

HAL has now support for Macbook Pro and Macbook laptop panels. Try hal >= 0.5.9rc3.

(In reply to comment #1)
> HAL has now support for Macbook Pro and Macbook laptop panels. Try hal >=
> 0.5.9rc3.
> 

Well there is a strange bug, (with both 0.5.9rc3 an 0.5.9)
it correctly set the backlight device but makes the macbook crash at the second suspend:
system starts : backlight correctly set
system suspends: the backlight is at maximum power and macbook-backlight returns
unexpected maximum brightness value
  please email desrt@desrt.ca about this. 
system suspends and crashs

However if i use fixbacklight after the first suspend the backlight function return to its normal state and the system can now correctly suspend without crashing

fixbacklight is a little hack which was done by Ryan Lortie, i think this little c code was commited in x because x now correctly wakeup after a suspend (xorg from git)

i still have the C code here :

//start

/*
 * Macbook Backlight Control
 * Copyright © 2006 Ryan Lortie <desrt@desrt.ca>
 * 
 * HAL integration Copyright © 2007 Ryan Lortie <desrt@desrt.ca>
 * using code Copyright © 2006 David Zeuthen <david@fubar.dk>
 *
 *   This program is free software; you can redistribute it and/or modify
 *   it under the terms of version 2 of the GNU General Public License as
 *   published by the Free Software Foundation.
 *
 *   This program is distributed in the hope that it will be useful,
 *   but WITHOUT ANY WARRANTY; without even the implied warranty of
 *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 *   GNU General Public License for more details.
 *
 *   You should have received a copy of the GNU General Public License
 *   along with this program; if not, write to the Free Software
 *   Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02110 USA
 *
 * This program was written after I reverse engineered the
 * AppleIntelIntegratedFramebuffer.kext kernel extension in Mac OS X and
 * played with the register at the memory location I found therein.
 *
 * From my experiments, the register appears to have two halves.
 *
 * yyyyyyyyyyyyyyy0xxxxxxxxxxxxxxx0
 *
 * The top (y) bits appear to be the maximum brightness level and the
 * bottom (x) bits are the current brightness level.  0s are always 0.
 * The brightness level is, therefore, x/y.
 *
 * As my Macbook boots, y is set to 0x94 and x is set to 0x1f.  Going below
 * 0x1f produces odd results.  For example, if you come from above, the
 * backlight will completely turn off at 0x12 (18).  Coming from below,
 * however, you need to get to 0x15 (21) before the backlight comes back on.
 *
 * Since there is no clear cut boundry, I assume that this value specifies
 * a raw voltage.  Also, it appears that the bootup value of 0x1f corresponds
 * to the lowest level that Mac OS X will set the backlight I choose this
 * value as a minimum.
 *
 * For the maximum I do not let the value exceed the value in the upper 15
 * bits.
 *
 * Turning the backlight off entirely is not supported (as this is supported
 * by the kernel itself).  This utility is only for setting the brightness
 * of the backlight when it is enabled.
 */

#include <pci/pci.h>

#include <stdio.h>
#include <sys/mman.h>
#include <fcntl.h>
#include <unistd.h>
#include <stdlib.h>

enum { FALSE, TRUE };

#define REGISTER_OFFSET       0x00061254
#define PAGE_SIZE             4096
#define PAGE_MASK             (PAGE_SIZE - 1)

#define ACCESS_OFFSET         (REGISTER_OFFSET & PAGE_MASK)
#define ACCESS_INDEX          (ACCESS_OFFSET >> 2)

unsigned int *register_page;

static unsigned long
determine_video_base_address (void)
{
  struct pci_access *pacc;
  struct pci_dev *pdev;
  unsigned long address;
  int i;

  address = 0;

  pacc = pci_alloc ();
  pci_init (pacc);
  pci_scan_bus (pacc);

  for (pdev = pacc->devices; pdev; pdev = pdev->next)
  {
    pci_fill_info (pdev, PCI_FILL_IDENT | PCI_FILL_BASES);

    if (pdev->vendor_id == 0x8086 && pdev->device_id == 0x27a2)
      for (i = 0; i < 6; i++)
      {
        if (pdev->size[i] == 512 * 1024)
        {
          address = pdev->base_addr[i];
          goto end;
        }
      }
  }

end:
  pci_cleanup (pacc);

  return address;
}

static unsigned long
register_get (void)
{
  return register_page[ACCESS_INDEX];
}

static void
register_set (unsigned long value)
{
  register_page[ACCESS_INDEX] = value;
}


static int
map_register_page (void)
{
  long address;
  int fd;

  address = determine_video_base_address ();

  if (address == 0)
    return FALSE;
  
  fd = open ("/dev/mem", O_RDWR);

  if (fd < 0)
    return FALSE;

  register_page = mmap (NULL, PAGE_SIZE, PROT_READ | PROT_WRITE,
                        MAP_SHARED, fd,
                        (address + REGISTER_OFFSET) & ~PAGE_MASK);

  close (fd);

  if (register_page == MAP_FAILED)
    return FALSE;

  return TRUE;
}

int
main (void)
{
  if (!map_register_page ())
    fprintf (stderr, "cannot map register page!\n");
  else
    register_set ((0x94 << 17) | (0x94 << 1));

  return 0;
}

//end

Eloi Primaux,
Do you still experience this issue with newer soft ?
Please check the status of your issue.
Or close this bug.