Jump to content
Welcome to InsanelyMac Forum

Register now to gain access to all of our features. Once registered and logged in, you will be able to contribute to this site by submitting your own content or replying to existing content. You'll be able to customize your profile, receive reputation points as a reward for submitting content, while also communicating with other members via your own private inbox, plus much more! This message will be removed once you have signed in.

Stratus Fear

Members
  • Content count

    102
  • Joined

  • Last visited

About Stratus Fear

  • Rank
    InsanelyMac Geek

Profile Information

  • Gender
    Male
  1. So I did notice that if I put a large amount of data through the card for ten minutes or so (say, Time Machine, for example) it simply stops sending and receiving data until I unload and reload the kext. I don't know if this issue is particular to my model or what. Nothing seems to be output in the logs about it, either. Hmm...
  2. The only other thing I noticed was a couple of register definitions that were using reserved bits, and one for powering down the device was using a bit that would require a PCI bus reset and not just a software command to power back up (PHY IDDQ, if my memory serves, it was off by only one bit, 20 vs 21 i think). If I was wrong in changing those, of course revert those changes I also notice someone posted an issue in your repo re: ifconfig en0 down and ifconfig en0 up. When I do the first command, my machine actually KPs for some reason. I'll add the note to that issue.
  3. Just for the record here are the diffs that came out of github:mac: BCM5722D.cpp: @@ -36,6 +36,7 @@ static const struct SupportedDevice { DEVICEID_BCM5787M, "BCM5787M NetLink (tm) Gigabit Ethernet" }, { DEVICEID_BCM5906, "BCM5906 NetLink (tm) Fast Ethernet" }, { DEVICEID_BCM5906M, "BCM5906M NetLink (tm) Fast Ethernet" }, + { DEVICEID_BCM57788, "BCM57788 NetLink (tm) Gigabit Ethernet" }, { 0, NULL } }; Info.plist: @@ -39,6 +39,7 @@ <string>pci14e4,1672</string> <string>pci14e4,167b</string> <string>pci14e4,1673</string> + <string>pci14e4,1691</string> <string>pci14e4,169b</string> <string>pci14e4,1693</string> <string>pci14e4,1712</string> PHY.cpp: @@ -687,7 +687,7 @@ void BCM5722D::resolveOperatingSpeedAndLinkDuplex(UInt16 status) media.speed = kLinkSpeed1000; media.duplex = kLinkDuplexFull; - currentMediumIndex = kMediumTypeIndex1000HD; + currentMediumIndex = kMediumTypeIndex1000FD; break; Register.h: @@ -30,6 +30,7 @@ #define DEVICEID_BCM5787M 0x1693 #define DEVICEID_BCM5906 0x1712 #define DEVICEID_BCM5906M 0x1713 +#define DEVICEID_BCM57788 0x1691 #define PHYID_MODEL_5754_5787 0x0E #define PHYID_MODEL_5755 0x0C @@ -729,17 +730,17 @@ enum { #define PHY_AUTONEGADVERT_802_3 BIT(0) #define PHY_AUTONEGADVERT_10HD BIT(5) #define PHY_AUTONEGADVERT_10FD BIT(6) -#define PHY_AUTONEGADVERT_100FD BIT(7) -#define PHY_AUTONEGADVERT_100HD BIT(8) +#define PHY_AUTONEGADVERT_100HD BIT(7) +#define PHY_AUTONEGADVERT_100FD BIT(8) #define PHY_AUTONEGADVERT_PAUSECAP BIT(10) #define PHY_AUTONEGADVERT_ASYMPAUSE BIT(11) // Auto-Negotiation Link Partner Ability Register #define PHY_AUTONEGPARTNER 0x05 -#define PHY_AUTONEGPARTNER_10FD BIT(5) -#define PHY_AUTONEGPARTNER_10HD BIT(6) -#define PHY_AUTONEGPARTNER_100FD BIT(7) -#define PHY_AUTONEGPARTNER_100HD BIT(8) +#define PHY_AUTONEGPARTNER_10HD BIT(5) +#define PHY_AUTONEGPARTNER_10FD BIT(6) +#define PHY_AUTONEGPARTNER_100HD BIT(7) +#define PHY_AUTONEGPARTNER_100FD BIT(8) #define PHY_AUTONEGPARTNER_PAUSECAP BIT(10) #define PHY_AUTONEGPARTNER_ASYMPAUSE BIT(11) @@ -750,8 +751,8 @@ enum { // 1000BASE-T Status Register #define PHY_1000BASETSTS 0x0A -#define PHY_1000BASETSTS_PARTNERFD BIT(10) -#define PHY_1000BASETSTS_PARTNERHD BIT(11) +#define PHY_1000BASETSTS_PARTNERHD BIT(10) +#define PHY_1000BASETSTS_PARTNERFD BIT(11) #define PHY_DSPRWPORT 0x15
  4. In trying out the driver, I noticed that, no matter what I did, I could not get a 1000FD connection. Even if I forced it and my switch is advertising it, the PHY would only set to half-duplex. I would see this in the output of ifconfig: media: autoselect (1000baseT <half-duplex>) When forcing: media: 1000baseT <full-duplex> (1000baseT <half-duplex>) I took at look at the code and found spots in Register.h that look to have accidentally had HD and FD bits swapped according to the BCM5722 Programmer's Guide (I found some in 10, 100, and 1000 modes for advertisement and negotiation, as well as mode set), and the section of PHY.cpp that sets the PHY mode was setting 1000HD register bits for both 1000HD and 1000FD mode. I fixed those and added my network controller to the model list (0x1691). Upon testing the modifications, the card is recognized and model name is output in the system logs, and 1000FD appears to be set correctly. media: autoselect (1000baseT <full-duplex>) My switch also reports 1000baseT FDX operation. I attached the modified files below in a zip. Please feel free to double check any changes I made. I'm not really a programmer, but I know enough to be able to follow what's happening in source code, so hopefully I didn't break anything. But it appears to work correctly. Thanks so much for creating this driver! I was ripping my hair out trying to get my card to work in Lion with AppleBCM5701Ethernet to no avail, as the previously hex edited driver that worked in SL had unresolved symbol linking errors in Lion. Archive.zip
  5. Works great on 10.7 64-bit with a BCM57788 (14e4:1691). Thanks!
  6. Leopard build 9A283 has leaked!

    Alternatively, you may be able to put this in the app package's specific frameworks folder. It'd be at iMovie HD.app/Contents/Frameworks. Give that a try before replacing system files -- it might make iMovie work, and keep the rest of your programs running off of the newer system framework.
  7. Yes, it is streched. Windows doesn't have a driver to control the LED. The LED isn't lit in Mac OS normally either unless the machine is asleep or the screen is off. For Windows, the LED only lights up during standby, not even when the LCD is off. Why the hell would anyone call that fake? That's just retarded. It's not hard to BSOD Windows. In the past it wasn't hard to kernel panic Mac OS either, if you knew what bugs to exploit (i.e., copy a folder inside of itself -- fixed in a certain Jaguar point release, and raiding dmg files in disk utility, also fixed at some point).
  8. The Great Driver Hunt

    There are other versions of the broadcom driver on this page: http://www.station-drivers.com/page/broadcom.htm You can try to see if any of them help your wireless reception.
  9. The Great Driver Hunt

    Use these drivers for the wireless. http://www.station-drivers.com/telechargem...%20bcm-43xx.exe Works perfectly for me. You might need to go to Device Manager at that device and use "Update Driver..." though after you install the drivers. They should work though.
  10. 20" iMac functionality

    Everyone using Pro on an iMac 20" has been getting that. Home supposedly works though.
  11. XP on an imac, up and booting (no joke this time)

    No. I've got a 20" iMac, and all it does is spit out the XOM console, can't get XP setup to ever display (even though you can tell setup is loading in the background -- noisy CD-ROM drive)
  12. The Great Driver Hunt

    You shouldn't do that. If you do, the machine won't boot with video again. What we need to do is emulate a VGA BIOS within the XOM bootloader. It seems like narf and blanka used a linear framebuffer as opposed to a real VGA driver? not sure. So, if they do that, then the Radeon drivers should work...
  13. The Great Driver Hunt

    Wonder if the device ID is different or something... The bluetooth chipset is indeed the same as the old D-Link USB bluetooth adapter I used to use, and those drivers worked on it back then... So it seems as though they would need to emulate a VGA BIOS to get around using the linear framebuffer deal... Hmm... As it is, it seems like they aren't. But I don't know about the internals of the code, so...
  14. The Great Driver Hunt

    Here's bluetooth, hope it works: http://www.megaupload.com/?d=0TI5FLDN
  15. The Great Driver Hunt

    So I've got some bluetooth drivers that I think should work, the Broadcom Widcomm ones. I pre-patched them so they don't have the not-the-right-OEM issue. Putting it up on Megaupload, but since it's going to take about 40 minutes to upload, and I've got an exam this morning, I'll have to post the link after I wake up. Sorry bout that...
×