anything in bfs_close() does not really makes sense. Thanks to Ingo for
pointing this out.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7429 a95241bf-73f2-0310-859d-f6bbb57e9c96
api version will be set to 3 as required for Dano/Zeta - apparently, the
previous mechanism doesn't work with our headers, as we don't define
B_BEOS_VERSION_DANO at all :)
Fixed the build again (because of the stdio switch).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7428 a95241bf-73f2-0310-859d-f6bbb57e9c96
NET_STACK_CONTROL_MODULE IPC support is still a work in progress, thought.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7419 a95241bf-73f2-0310-859d-f6bbb57e9c96
of "obfs" (thanks Ingo). Also, changed the destination directory. It was formelly
"beos/system/add-ons/kernel/file_systems/bfs/" and now it is
"beos/system/add-ons/kernel/file_systems/" (as it should be).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7411 a95241bf-73f2-0310-859d-f6bbb57e9c96
Fixed a possible endian problem in CheckBlockRun(), too.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7405 a95241bf-73f2-0310-859d-f6bbb57e9c96
BFS replacement (will report itself as "bfs" instead of "obfs". Is there a
way to change the add-on name based on the same define?
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7403 a95241bf-73f2-0310-859d-f6bbb57e9c96
too late, and could theoretically already be acquired by someone else).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7402 a95241bf-73f2-0310-859d-f6bbb57e9c96
FreeBSD compatibility layer that provides a struct ifnet interface,
all #defines that are used by the driver, the mbuf functions,
mbuf queues, etc.
Also some unused things are provided by empty macros.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7399 a95241bf-73f2-0310-859d-f6bbb57e9c96
from within an interrupt handler, and wastes pretty much memory.
Knows only about two different block sizes, and implements them by
using a simple free list.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7398 a95241bf-73f2-0310-859d-f6bbb57e9c96
util.* provide a convenience function to map physical memory.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7397 a95241bf-73f2-0310-859d-f6bbb57e9c96
Initialized attaches and detaches the core driver, sends and receives packets
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7395 a95241bf-73f2-0310-859d-f6bbb57e9c96
as this involves some long delays like eeprom reading, etc.
Need to snooze() there, and this moved this into a separate thread.
Obviously, the FreeBSD driver isn't doing prober locking yet, so
it's not really save. But it's the correct way to do it, and it
doesn't crash. Inside the interrupt handler it did crash when reconnecting the cable...
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7393 a95241bf-73f2-0310-859d-f6bbb57e9c96
PCI id table, em_probe() and em_shutdown() are now gone.
Changed interrupt handler to use the BeOS calling convention, this avoid putting a wrapper around it.
em_attach() and em_detach() no longer link the adapter struct into a list.
When PCI io port access is configured, make sure that io address decoding is enabled in the PCI configuration.
replaced some offsetof() macros with #defiens to avoid importing all these structs.
Moved sysctl stuff into a #if 0 #endif wrapper.
Removed #include of FreeBSD header files.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7390 a95241bf-73f2-0310-859d-f6bbb57e9c96
Realtek RTL8169 Family Gigabit Ethernet Driver
Very basic, does only detection of the card and sets up some buffer descriptors.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7380 a95241bf-73f2-0310-859d-f6bbb57e9c96
These changes have been made by YellowTab and have been contributed to OpenBeOS.
While it might not fix every issue, things should be much better and more stable now.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7379 a95241bf-73f2-0310-859d-f6bbb57e9c96
Released under the BSD license.
===========================================
From: "Richard Houle" <richard.houle@sand.com>
Subject: RE: License question regarding beos via rhine driver.
Date: Thu, 1 Apr 2004 08:48:02 -0500
Whoa. You talking about very old stuff. I don't even know if I kept a copy
of that source code... If you have a copy of it, then you can do what you
want. I don't have any objections or restrictions.
For the license question, the driver was directly inspired by the FreeBSD
driver's source code. Therefore, It's a BSD License.
Richard Houle
===========================================
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7378 a95241bf-73f2-0310-859d-f6bbb57e9c96
UDP did not set AF_INET in the sockaddr that is passed to recvfrom().
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@7211 a95241bf-73f2-0310-859d-f6bbb57e9c96
