summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2012-10-26cleanup oce_init_fw; use less bitfieldsMike Belopuhov
2012-10-26rename oce_config_nic_rss to oce_config_rss and merge oce_rss_itbl_init in;Mike Belopuhov
cleanup oce_set_promisc and oce_config_vlan function arguments, a bunch of defines and prototypes while at it.
2012-10-26Previous revision would prevent grtwo(4) boards from being recognized.Miod Vallat
Make sure that what seems to be a 8-bit ID register of value zero does not match as an expansion board, which it can't be, and continue with the specific frame buffer test logic.
2012-10-26Properly clear trap frame in setregs() to avoid leaking registers across exec.Mark Kettenis
ok jsing@
2012-10-25make oce_mbox_{dispatch,wait} more readable and use less bitfieldsMike Belopuhov
2012-10-25simplify/unify writes to the rx and tx doorbell registersMike Belopuhov
2012-10-25don't use bitfields in oce_pci_alloc; kill some unused structuresMike Belopuhov
2012-10-25cleanup interrupt register definesMike Belopuhov
2012-10-25Fix a hilarious bug where oce_set_promisc wouldn't zero out theMike Belopuhov
command buffer it sent to the card but an incorrect assignment of the DMA address pointing to this buffer would offset the problem. Once I've fixed an incorrect assignment oce_init started to fail. Turns out that oce_set_promisc is about the only function whose command size is larger than the payload of the standard command buffer and needs an external chunk of DMA memory.
2012-10-25simplify oce_arm_eq and oce_arm_cqMike Belopuhov
2012-10-25Some LIBC_SCCS found lurking in the shadowsTheo de Raadt
2012-10-23Only clear the revarp_ifp if the ifp pointing to it is detached.Claudio Jeker
Since the ifp is not detached in the SIOCSIFRDOMAIN case remove the clear there. With and OK deraadt@
2012-10-22Stop passing around PS_STRINGS in %g1. The ELF ABI reserves this registerMark Kettenis
for passing around a pointer to a cleanup function and we'd like to use it for that purpose in the near future. ok miod@
2012-10-22Stop passing around PS_STRINGS in %g1. The ELF ABI reserves this registerMark Kettenis
for passing around a pointer to a cleanup function and we'd like to use it for that purpose in the near future. ok miod@
2012-10-22- Provide flow control statusBrad Smith
- Use ANAR_PAUSE_TOWARDS define - Only advertise flow control if the MAC sets MIIF_DOPAUSE ok jsg@
2012-10-22Add flow control support.Brad Smith
ok jsg@
2012-10-22Provide flow control status.Brad Smith
ok jsg@
2012-10-22Add flow control support.Brad Smith
Tested with Apple/Sun gem(4) with and without flow control (Sun ERI/Apple GEM with FastE PHY). ok kettenis@
2012-10-22missing headerMiod Vallat
2012-10-22Rewrite the receive filter handling code and cleanup the ioctl bits.Brad Smith
Tested by and ok stsp@
2012-10-22Fix the multicast filter full size check.Brad Smith
ok mikeb@
2012-10-21Fix thinko in previous.Mark Kettenis
2012-10-21Fix problem reported by Nathan Weeks <weeks@iastate.edu> where a userlandBob Beck
program could induce the kernel to panic by attempting to do a sempo with nsops > kern.seminfo.semume and the SEM_UNFO flag set. This fixes it so we return ENOSPC, like the man page says, rather than panicing. ok miod@, millert@
2012-10-21Make sure that there is enough room in the transmit queue to send a VIOMark Kettenis
message. Sleep if necessary.
2012-10-21I want msdos filesystem support on the big bsd.rdTheo de Raadt
2012-10-21Clear a pending channel reset when a device is opened.Mark Kettenis
2012-10-21repair pci powerstate handling. The extra timeouts were not needed.Theo de Raadt
Furthermore, all the extra powering up and down caused resumes to not work. ok ratchov
2012-10-21Add the IP_DIVERTFL socket option on divert(4) sockets to controlSebastian Benoit
which packets (as in direction) of the traffic will be diverted through the divert socket. ok claudio@, henning@
2012-10-21Needs <sys/device.h> nowMiod Vallat
2012-10-20driver xxactivate() functins are run even if a driver xxattach() hasTheo de Raadt
"failed"; no indication is given up to the autoconf framework since xxattach() returns void. Until this situation is improved, there are a handful of drivers which must use driver-specific checks in xxactivate() to see if the xxattach() function succeeded or failed. Add such a check here. ok kettenis
2012-10-20move sio_intr_shutdown() from being a shutdownhook, and do it directly inTheo de Raadt
boot() as required. ok miod
2012-10-20Implement transmit DMA segments. The descriptor usage is from FreeBSD,Chris Cappuccio
which contradicts the datasheet (but actually works). Also, use a segment (instead of a new mbuf and data copy) to zero-pad small packets, modeled after if_myx. ok sthen@, gerhard@, henning@
2012-10-20off-by-one in device number check; found the hard way and reported by J Sisson.Miod Vallat
2012-10-20Init values for the AR9485 were for version 1.0 of this chip, which accordingStefan Sperling
to Atheros Linux developers was never sold. So update initvals to what Linux is using for the 1.1 generation. Because the serdes values are written to different registers on the AR9485 this involves tweaking the serdes init code for all athn(4) chip families. This commit doesn't make AR9485 devices work yet but is a step in the right direction. Tested by krw, kettenis, and Andrew Ngo. ok kettenis@
2012-10-20Fix hardware kill switch detection for the ar9300 chip family. The driver wasStefan Sperling
checking the wrong bits of eeprom to determine rfkill switch pin and polarity, and was reading the wrong register to determine rfkill gpio pin state. ok kettenis@
2012-10-19add an #ifdef HIBERNATE for diffability with amd64Mike Larkin
2012-10-19amd64 hibernate "unpack-time" mmu/pmap code and asm goo. Work inMike Larkin
progress.
2012-10-19andre.s@list.ru has a laptop which does not bring the azalia out ofTheo de Raadt
D3 state, so the attach funciton must do this. ok kettenis
2012-10-19small cleanups and output prettyingTheo de Raadt
ok yuo matthieu
2012-10-19Be sure to compile boot block objects with -D_NO_ABICALLS, to prevent .S filesMiod Vallat
in libkern from getting a ".abicalls" pseudo-op from including <machine/asm.h>; this will in turn flag the .o file as PIC even though it doesn't contain PIC code, really. This gets rid of the annoying "linking PIC files with non-PIC files" warnings spewed by ld at link time.
2012-10-19Link boot64 twice, first as a self-contained binary, to enforce there are noMiod Vallat
unresolved symbols, then as a relocatable image as initially intended. This will prevent the arcbios.c 1.18 breakage from occuring again.
2012-10-19Simplify xl_iff_90x() a bit and only check ac->ac_multicnt since ac_multicntBrad Smith
is also bumped for multicast ranges. ok mikeb@
2012-10-18Convert a number of old private copies of code which predatesTheo de Raadt
pci_set_powerstate() to using it instead. Many of these chunks of code had bugs in them, especially missing delay() calls. Some of them were doing things our PCI subsystem is now responsible for handling. If you have any of the affected devices, please keep an eye out for regressions. ok kettenis
2012-10-18Use pci_set_powerstate() unconditionally for all chip variants, ratherTheo de Raadt
than a kludged together chunk of code for just one chip, which happens to be entirely wrong. We need to watch for regressions on cardbus, though. ok kettenis
2012-10-18Switch luna88k to timecounters; tested by aoyama@Miod Vallat
2012-10-18Put back the kl_n_shift variable, removed by mistake in 1.18, causing theMiod Vallat
boot blocks to fail on IP27 and IP35 (IP26, IP30 and 32-bit ARCBios systems unaffected). While there, provide a simpler ARC_Call() macro for 32-bit boot blocks, and put a few more variables and statements within explicit __LP64__ guards. Crank boot blocks version as well.
2012-10-18simplify checkreplaywindow() API; make call/return code handling consistentMarkus Friedl
ok mikeb@
2012-10-18make oce_arm_{eq,cq} functions look like the other queueMike Belopuhov
manupulation functions in this file; do some minor style cleanup while here.
2012-10-18Wiring map entries with VM_PROT_NONE only waists RAM and bears noGerhard Roth
advantages. We shouln't do this. If the protection changes later on (and VM_MAP_WIREFUTURE was set), uvm_map_protect() will wire them. Found by Matthias Pitzl. ok miod@ markus@
2012-10-18Move revarp code into #ifdef NFSCLIENT; saving space on some media.Theo de Raadt
Resurrect the rather silly "unplug my network device while I am doing nfs diskless revarp" safety code which was disabled due to a missing "ether.h" include, found by jsg ok jsg
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-17 19:11:56 +0000