summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2005-11-25update.Brad Smith
2005-11-25fix BCM5714 PCI id name and add 5714S, 5715 and 5715S.Brad Smith
2005-11-25regenBrad Smith
2005-11-25- add BCM5714S, BCM5715 and BCM5715SBrad Smith
- change BCM5714C to BCM5714 From Broadcom's Linux driver
2005-11-25Use BGE_IS_5705_OR_BEYOND macro to check for the existence or non-existenceBrad Smith
of 5705 or derived ASICs and remove the BGE_QUIRK_5705_CORE flag.
2005-11-25splimp -> splhigh in setroot()/rootconf().Brad Smith
ok miod@ drahn@
2005-11-25mention BCM5715Brad Smith
2005-11-25remove prototype for a non-existent function bge_free_jumbo_mem.Brad Smith
2005-11-25Use BGE_IS_JUMBO_CAPABLE in one more spot.Brad Smith
2005-11-25load the 5750 DSP code on the 5705 PHY too.Brad Smith
Based on Broadcom's Linux driver.
2005-11-25use 5780 here instead of 5714.Brad Smith
2005-11-25don't bother setting error in bge_attach() when its not actuallyBrad Smith
being used for anything.
2005-11-25Use BGE_IS_575X_PLUS macro when checking whether to use the PCI Express test ↵Brad Smith
or not.
2005-11-25Use BGE_IS_JUMBO_CAPABLE macro to figure out whether weBrad Smith
want Jumbos or not. Instead of just assuming that all 5705 derived cores do not support Jumbos. All of Broadcom's new Gig chips do not support Jumbos with the exception of the BCM5714 found embedded in the ServerWorks HT-2000 chipset using the BCM5780 ASIC.
2005-11-24add a few macros allowing selection of certain generations ofBrad Smith
the Broadcom ASICs.
2005-11-24Prefer vaddr_t and paddr_t types in device softc, instead of void * andMiod Vallat
heavy casts. Improves readability, no functional change.
2005-11-24Peek at the correct location in badpaddr() - how could this go unnoticed before?Miod Vallat
2005-11-24Remove unnecessary cast in an atop() construct.Miod Vallat
2005-11-24dont need spls in attachDavid Gwynne
2005-11-24enable esmDavid Gwynne
ok deraadt@
2005-11-24o Force the sending of an extra URB if there is less than 2 bytes leftDamien Bergamini
at the end of the last URB (URB = USB Request Block = 64 bytes). o Better check of the size of received frames. o In ural_tx_{data,mgt} send the packet to BPF after setting the duration field of the frame. o Minor tweaks.
2005-11-24remove __GNUC_PREREQ__(2, 96) block because lint pukes; idea for solution ↵Theo de Raadt
from miod
2005-11-24use WANTLINT= (on all architectures)Theo de Raadt
2005-11-24syncTheo de Raadt
2005-11-24add lint-specific hacks. at the same time, clean out a lot of ancientTheo de Raadt
cruft for old compilers and environments. there may be a 2nd round of polishing after this; ok miod and others
2005-11-24fix use after free introduced in r1.30. ok deraadt@Moritz Jodeit
2005-11-24bye bye lkm/vfs dirTheo de Raadt
2005-11-24more empty files for lint to chew on (on the vax)Theo de Raadt
2005-11-24include more architectures in the list (even though some do not compileTheo de Raadt
this, but do it entirely in the compiler); ok miod
2005-11-24code is not being used; toss into AtticTheo de Raadt
2005-11-24memset();Xavier Santolaria
2005-11-24wi and ubt are network drivers so should be DV_IFNET,Alexander Yurchenko
not DV_DULL. Spotted by miod@.
2005-11-24"hmac" not "hmc", notice by <gwyllion at ulyssis dot org>Hans-Joerg Hoexer
2005-11-24ural should be DV_IFNET, not DV_DULL. noticed by ian@.Alexander Yurchenko
ok many.
2005-11-24don't redefine BGE_PCIMISCCTL_ENDIAN_WORDSWAP and lower the delay whileFederico G. Schwindt
waiting for the firmware. tested and ok brad@
2005-11-24Remove kernfs, okay deraadt@.Pedro Martelletto
2005-11-24ike has used hmac sha1/md5 since ever. So call 'em hmac-sha1/md5.Hans-Joerg Hoexer
2005-11-24remove tests for keyed md5/sha1, we only have hmacsHans-Joerg Hoexer
2005-11-24Make clear we only have "hmac-sha1" and "hmac-md5".Hans-Joerg Hoexer
2005-11-24Remove old-style keyed sha1/md5. We only support hmac-sha1/md5.Hans-Joerg Hoexer
Noticed the hard way by <raff at brodewicz dot pl>
2005-11-24respect TMPDIR;Xavier Santolaria
2005-11-24minor knf + snprintf() return check;Xavier Santolaria
2005-11-24typo; pr4916; from frantisek holop <minusf@obiit.org>Michael Shalayeff
2005-11-24#if 0 doesnt belong hereDavid Gwynne
2005-11-24kill trailing whitespace;Jason McIntyre
2005-11-24missing full stop;Jason McIntyre
2005-11-24repair indent that cvs has scrwed make it extra confusingMichael Shalayeff
2005-11-24print something out when we hit unsupported devices. hopefully we canDavid Gwynne
pick up a dmesg with them.
2005-11-24white space fixesDavid Gwynne
2005-11-24break the 1:1 mapping of esm sensors to kernel sensors. according toDavid Gwynne
jordan theres an esm sensor that packs the status of four drive slots into a single reading, but it should be represented in the kernel as four separate sensors. some of the sensors (such as the pci slot ones and the hotplug status) should probably be dealt with in the same way. so esm sensors now have a list of their own types, and a mapping of these types to kernel sensor types. instead of storing the kernel sensor in the esm_sensor struct, allocate the two separately so esm_sensor can point to multiple kernel sensors. make the drives sensor fill out 4 kernel sensors. it still works on my box, but i dont have a machine old enough to see if the packed esm sensor turns into 4 kernel ones without problems. i hate how theres no mechanism to remove kernel sensors after you use SENSOR_ADD. ok marco@
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-28 05:46:30 +0000