|
|
ru.unix.bsd- RU.UNIX.BSD ------------------------------------------------------------------ From : Andrey Slusar 2:467/126 07 Jun 2006 22:11:41 To : All Subject : ЛПНХ ЬФП ЙОФЕТЕУОП... -------------------------------------------------------------------------------- --8<---------------cut here---------------start------------->8--- One other thing I've noticed is that system interactivity suffers a lot when the interfaces are heavily loaded, i.e. handling high frame rates. The interfaces keep going, but the shell becomes unresponsive until the traffic subsides. RealTek asked for driver support for 5.x, 6.x and 7.x. I think I have 6.x under control. Now I need to do 7.x. I put a copy of the 6.x code I'm testing now at: http://www.freebsd.org/~wpaul/RealTek/re.tar.gz The main changes are: - Added support for all the new PCI IDs and hwrevs in all the devices RealTek sent me (8169SB/8110SB, 8169SC/8110SC, 8168B/8111B, 8101E). - Finally fixed EEPROM access for all devices. (You have to be careful how you enable access to the EEPROM chip: first you flip the 'EEPROM programming mode' bit, and *then* you flip the EE chip select bit. Doing them both in one access doesn't work.) The station address is always read from the EEPROM now. - Modified the interrupt handling to use a 'fast' handler, similar to what I did in the VxWorks driver. This seems necessary for the PCIe devices, which appear to require that all their pending interrupt events be 'drained' otherwise they stop delivering interrupts entirely. - Modified the rgephy driver to only perform the DSP fixup for rev 0 and rev 1 of the on-board RealTek PHY in the gigE chips, per the RealTek-supplied code. --8<---------------cut here---------------end--------------->8--- -- Regards, Andrey. --- Gnus/5.1007 (Gnus v5.10.7) XEmacs/21.5-b27 (i386--freebsd) * Origin: Santinel (2:467/126) Вернуться к списку тем, сортированных по: возрастание даты уменьшение даты тема автор
Архивное /ru.unix.bsd/274910d64eed.html, оценка из 5, голосов 10
|