June 29 6:45 PM:

Current status: Up until now, I have completed most of the data structure definitions, skge_probe(), and skge_open(). I am obviously quite behind compared to my project plan, but that is not something I can easily resolve; However; It does look like I will have time still left over at the end of writing the skge driver for some other prospects in this task. Possibly tg3?

So, after basicly 3 days off (except for a little work on saturday during and after my meeting) I'm ready to get going. It's 6:47 PM and I'm about to start on skge_transmit().

June 29 2:00 AM (really June 30):

It's been a productive night; I discovered and fixed another memory issue with the allocation of Adapter; Once that was dealt with, I implemented skge_transmit and skge_poll. I'm not certian as I haven't setup wireshark yet, but I believe that I just saw my skge driver transmit it's first packet.

I also pretty much overhaulled the DBGP commands so that anyone interested in writing a driver can literally just look at the output and understand all that is going on.

skge_probe - start
|
|-> Allocating 88 bytes of private space
|-> Registering skge_operations with the core
|-> Creating link to device
|-> Allocated 68 bytes for the adapter at 0x18d74
|-> Zero out from 0x18d74 to 0x19f84
|-> Adapter Base set to 18d7c
|-> ioaddr set to d800
|-> Using IRQ Number 10
|-> Assigned the Back and pdev pointers
|-> Defining the length of the regiser locations
|-> Mac Address of Port 1 : 00:21:91:91:10:6d
|-> Beginning skge_initialize
|   |-> Triggering a software reset
|   |-> Waiting for one second . . . Done.
|   |-> Enabling Test Mode
|   |   |-> Removing error bits
|   |   |-> Disabling Test Mode
|   |
|   |-> Chip ID:
|   |   |-> Detected   : b1
|   |   |-> Reference Examples:
|   |       |-> Genesis    : a
|   |       |-> Yukon      : b0
|   |       |-> Yukon Lite : b1
|   |       |-> Yukon LP   : b2
|   |
|   |-> Detecting Chip Type:
|   |   |-> Marvell
|   |
|   |-> Configuration Register   : 0x93
|   |-> Number of Ports Detected : 1
|   |-> Chip Revision            : 0x9
|   |-> Calculating Ram Properties
|   |   |-> ram_size                 : 65536
|   |   |-> ram_offset               : 0
|   |
|   |-> Interface Mask is Register Value.
|   |-> Forking into card specific initialization for Non-Genesis settings
|   |   |-> Switching to VCC
|   |   |-> Checking for stuck hardware bit (not-stuck (good))
|   |   |-> Clearing error bits
|   |   |-> Performing Software Reset
|   |
|   |-> Stopping the card's hardware clocks
|   |-> Turning the LED on
|   |-> Enabling the transmit arbiter for port #0
|   |-> Setting timeout initial values
|   |-> Setting up the clocks
|   |-> Resetting port #0
|   |-> Entering yukon_reset
|   |   |-> Disabling PHY IRQ's
|   |   |-> Clearing MC Hash 1/4
|   |   |-> Clearing MC Hash 2/4
|   |   |-> Clearing MC Hash 3/4
|   |   |-> Clearing MC Hash 4/4
|   |   |-> Clearing temporary registers
|   |
|
|-> Entering skge_devinit: 0
|   |-> Linking to private space
|   |-> Setting up pointers
|   |-> Defining tx (8) and rx (8) ring sizes
|   |-> Defining modes, duplexing, and advertising
|   |-> Setting link down prior to actual link detection
|
|-> Registering Netdev
|-> Setting Pci Device pointer
|-> Exiting skge_probe (return:0)




gPXE 0.9.7+ -- Open Source Boot Firmware -- http://etherboot.org
Features: HTTP DNS TFTP AoE iSCSI bzImage COMBOOT ELF Multiboot PXE PXEXT

skge_open - start
|
|-> Defining Receive Buffer of 1536
yukon_mac_init - Yukon Lite >= A3 - Yukon Lite >= A3 - Autoneg enabled. - flow:S or R-
Function: yukon_init - start
Function: yukon_init - end
yukon_mac_init - endadapter : 101748
rxqaddr : 85956
port    : 0
ram_addr: 0
chunk: 32768
skge_open - end - return 0
net0: 00:21:91:91:10:6d on PCI00:08.0 (open)
  [Link:down, TX:0 TXE:0 RX:0 RXE:0]
Waiting for link-up on net0...skge : phy interrupt status 0x40 0x8100
skge : phy interrupt status 0x9040 0x514c
skge : phy interrupt status 0x5c60 0x7d2c
Yukon - link up - startskge net0: Link is up at 100 Mbps, full duplex
Yukon - link up - end ok
DHCP (net0 00:21:91:91:10:6d).skge_transmit - start
skge_transmit - end
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
skge : phy interrupt status 0x0 0x6d2c
.skge_transmit - start
skge_transmit - end
skge : phy interrupt status 0x0 0x6d2c

Up to july 3, big hole in my memory as to what happened when: Basicly, I spent some time working with Andytim trying to trace out what was happening with my transmit method and never ended up finding anything wrong. A few other eyes looked at it with me and … no results. – Chris