Differences

This shows you the differences between two versions of the page.

--- soc:2008:stefanha:journal:week4 [2008/06/19 08:58]
stefanha
+++ soc:2008:stefanha:journal:week4 [2008/06/23 04:58] (current)
stefanha
@@ Line 138: / Line 138: @@
 ==== Thu Jun 19 ====
+Git commit: [[http://git.etherboot.org/?p=people/stefanha/gpxe.git;a=commit;h=2c5b2a45b7c33b1cd45e99c29f0a833c458e2ecb|[b44] Create skeleton driver for Broadcom 4401 NIC]]
 **Brought up ROM-o-matic for Etherboot top-of-git-tree**.  I have been occasionally assisting mdc with his [[http://rom-o-matic.net/|ROM-o-matic.net]] online boot ROM generator.  He recently enabled ROM-o-matic for gPXE top-of-git-tree.  That way users can get ROMs for the latest development version of gPXE without having to set up a development environment and build from source.  This is now also possible for Etherboot.
 **Beginning work to port Linux b44 (Broadcom 4401) driver**.  My laptop has a BCM4401-B0 NIC and is currently not supported by gPXE.  The idea is to port the Linux driver to gPXE.  I am looking forward to learning more about network drivers and device driver development in general.
-Next steps:
+==== Fri Jun 20 ====
-  * Update [[:dev:gdbstub|GDB stub page]] and screencast when UDP code is merged into mainline.  See [[http://grub.enbug.org/DebuggingWithGDB|GRUB GDB wiki page]] for inspiration.
+Git commit: [[http://git.etherboot.org/?p=people/stefanha/gpxe.git;a=commit;h=598d23c5b0ec778dcbaafe634569ca9f8f8d1854|[b44] Minimal TX path]]
-  * gPXE bzImage support.
-  * Real-mode GDB stub.
+**The b44 driver is transmitting Ethernet frames**.  Thanks to Michael Decker's excellent [[:soc:2008:mdeck:notes:gpxe_driver_api|gPXE Driver API Documentation]] I got the skeleton for the driver working very quickly last night.  This morning I started porting the Linux b44 driver code.
+After getting the initialization working (mainly by copy-paste) and reading the MAC address from the card, I decided to pursue the TX path.  Getting transmit working early is useful since gPXE will attempt to do DHCP automatically and therefore needs to send packets.
+Copy-pasting the Linux driver was not a good tactic since the Linux code is much more complex.  Eventually I just focused on understanding how the hardware supports transmitting frames (there is no public documentation available!), and then implemented a simple TX path resembling the gPXE natsemi driver.
+==== Sat Jun 21 ====
+Git commit: [[http://git.etherboot.org/?p=people/stefanha/gpxe.git;a=commit;h=39144f971c261bcd92f5052059d444d742b0010f|[b44] Working RX path]]
+**The b44 driver is receiving Ethernet frames**.  I just booted PXELINUX and HTTP-booted Linux 2.6.25 on this card for the first time!  Getting the RX path working has been painful.
+I think some of the Linux driver code is misleading/incorrect.  Luckily there are drivers for OpenBSD, FreeBSD, and Solaris.  Those drivers might even be based on the Linux driver, but they do some things differently and it helps to compare them to each other.  My main issue with the RX path was a comment in the Linux driver claiming that the hardware writes a header structure 30 bytes //before// the DMA address of the I/O buffer.
+This is false.  The Linux driver does offset the DMA address by 30 bytes, but it also offsets the IO buffer by 30 bytes.  In the end, it makes no difference and all that has happened is that 30 bytes of the IO buffer have been wasted.  The header structure gets written //to// the DMA address, not before it.
+The next steps for the b44 driver are cleaning it up, making it robust, and testing.  Most of the initialization code is straight from the Linux driver.  I want to get to grips with it and then simplify it for gPXE.
+I have omitted performance optimizations from the Linux driver.  The Linux driver has a "copy threshold" which dictates whether to copy a received packet to a fresh IO buf to hand off to the network stack, or whether to remove the current IO buf from the RX ring and pass it straight to the network stack (and allocating a fresh IO buf for the RX ring).  I'll talk to Balaji about performance measurement since he's been optimizing his USB driver.
+**Lilo bzImage debugging still underway**.  I made a little bit of progress tonight by determining that the triple-fault happens in the call to ''install''.  I think that EIP goes crazy somewhere inside ''install'' and hence the triple fault.  I'm sure the issue triggers inside ''install'' since I've placed infinite loops before and after the call.  The loop after the call never happens.
+My current debugging cycle is by booting up Damn Small Linux in QEMU and copying over my latest ''gpxe.lkrn'', running ''lilo'', and rebooting into gPXE.  This is slow and frustrating.  I need to script it but my DSL install seems to be read-only.
+===== Next week =====
+On to [[.:week5|Week 5]].

Trace:

Differences

Navigation

Search

Toolbox

QR Code