summaryrefslogtreecommitdiff
path: root/sys
AgeCommit message (Collapse)Author
2017-07-19Fix previous commit; check the enter_ddb function pointer instead of theMark Kettenis
"accessops" pointer before calling it.
2017-07-19Improve the heuristic which selects 5GHz APs over 2GHz APs.Stefan Sperling
The previous code wasn't quite right: it didn't account for the fact that some drivers don't set ic_max_rssi, and it compared 5GHz APs to a threshold relative to the max RSSI, rather than comparing RSSI on 5GHz relative to RSSI on 2GHz. This heuristic is only used by SCANNALLBAND drivers: iwn(4), iwm(4), wpi(4) In the future the AP selection heuristic should be made more intelligent, e.g. it should take BSS load information in beacons into account. Another open problem is inconsistent representation of RSSI measurement values throughout our drivers and stack. Help is welcome! For now, this hopefully improves AP selection at busy airports. ok sthen@ deraadt@
2017-07-19Turn this into a panic since there's no way to recover from itMike Belopuhov
2017-07-19Add a handler for RNDIS status messagesMike Belopuhov
RNDIS status messages may indicate various conditions reported by the hardware, including link status events. They're not supposed to be universally treated as errors. Issue reported by Artturi Alm, ok armani
2017-07-19Avoid additional div0 in case both ntracks & nsectors passed inTheo de Raadt
via ioctl are 0. Noticed by bluhm ok bluhm millert
2017-07-19Add a enter_ddb() "accessop" to wsdisplay(4) to allow KMS drivers to bypassMark Kettenis
the modeset lock when entering ddb. This avoids triggering various asserts when the kernel panics while running X. ok deraadt@
2017-07-19Uninitialized variable can leak kernel memory.Theo de Raadt
Found by Ilja Van Sprundel ok kettenis
2017-07-19Fix the test condition weakened a recent changeMike Belopuhov
2017-07-19Avoid division by zero of vnd_secsizeTheo de Raadt
Found by Ilja Van Sprundel ok claudio kettenis
2017-07-19Factor out internal FIFO queue managerMike Belopuhov
HFSC internal queue becomes accessible via pf queueing ops. It will also select an alternative queue manager based on the queue spec. Discussed with and OK henning@ at d2k17 as a part of a larger diff.
2017-07-19Rework HFSC vs FQ-CoDel checksMike Belopuhov
The selection mechanism introduced in pf_ioctl.c -r1.316 suffers from being too ambiguous and lacks robustness. Instead of relying on composition of multiple flags in the queue specification, it's easier to identify the root class (if it exists) and derive all further checks from it.
2017-07-19more depends gc / yacc rules overhaulMarc Espie
okay millert@
2017-07-19If second xdr_string_encode() fails in bp_getfile() m_freem() m sinceClaudio Jeker
this mbuf was allocated by the first call. Fixes possible memory leak. Found by Ilja Van Sprundel OK bluhm@ deraadt@
2017-07-19The proper way to check for successful allocation of an mbuf cluster isClaudio Jeker
by checking M_EXT flag. m_freem() the mbuf in the errorcase to prevent a mbuf leak. MCLGET does not set the mbuf to NULL on failure. Found by Ilja Van Sprundel OK bluhm@ deraadt@
2017-07-19Check if mbuf cluster allocation succeeded and on failure free mbuf andClaudio Jeker
return ENOBUFS. Found by Ilja Van Sprundel OK bluhm@ deraadt@
2017-07-19Fix memory leak in iec_get(). If MCLGET fails free the mbuf as well.Claudio Jeker
Found by Ilja Van Sprundel OK bluhm@ deraadt@
2017-07-19Fix possible double free in the TX dma ring handling. If the allocation ofClaudio Jeker
the cluster fails in vte_init_tx_ring() the mbuf is removed but the slot in the ring still holds the pointer. A possible call to vte_stop() would then cause a double free. Again set pointer to NULL to prevent this. Found by Ilja Van Sprundel OK bluhm@
2017-07-19Fix double free in pppoe_dispatch_disc_pkt(). If m_pulldown() fails m isClaudio Jeker
freed but another m_freem call happens later. Set m to NULL if m_pulldown() fails like in all other cases of this function. Found by Ilja Van Sprundel OK bluhm@
2017-07-19Move KTRPOINT call up. The lenght variable i is getting aligned and soClaudio Jeker
uninitialised data can be dumped into the ktrace message. Found by Ilja Van Sprundel OK bluhm@
2017-07-18Both syslog(3) and syslogd(8) truncate the message at 8192 bytes.Alexander Bluhm
Do the same in sendsyslog(2) and document the behavior. reported by Ilja Van Sprundel; OK millert@ deraadt@
2017-07-18Prevent integer overflow in WSDISPLAYIO_LDFONT ioctl.Mark Kettenis
Issue found by Ilja van Sprundel. ok deraadt@
2017-07-18Remove stray whitespace.Mark Kettenis
2017-07-18Re-order members of struct selinfoStefan Fritsch
This reduces its size from 24 to 16 on 64bit archs. ok deraadt@
2017-07-18Add sdmmc support for the i.MX6UL variant of the i.MX6 series. ThisPatrick Wildt
variant is essentially the same in regards to sdmmc as the i.MX6SX variant, which is why the device tree specifies that version. The i.MX6SX has a few quirks. One of these is related to tuning, which we don't do, the other one to the CAP1 register, which we don't use, and the third is for high speed support, where the "support bit" is missing for HS200. Of all these, only the one where the CAP1 register spills into the CAP register seems relevant to us. With this I can mount and newfs on the SD card. ok kettenis@
2017-07-18Stop automatically regenerating gap.o (and hence linking a new kernel)Theo Buehler
whenever make is issued. Only do this when there are actual source changes below /sys. This restores the behavior prior to the KARL implementation. Problem noticed and patiently explained by mpi implementation with a lot of help from espie ok mpi, espie, deraadt
2017-07-18Add clock support for the i.MX6UL variant of the i.MX6 series. ThePatrick Wildt
device tree uses completely differently numbered indices compared to the i.MX6QDL variant. This means we have to add another clock list and use the right one depending on the compatible. Also, on the UL variant there is another mux in front of the perclk divider, so we have to cope with that as well. With feedback from and "looks reasonable" kettenis@
2017-07-18Fixup the printf string, spacingMike Belopuhov
2017-07-18soreserve() modifies `so_snd' and `so_rcv' so asserts that it is calledMartin Pieuchot
with the socket lock. This change is safe because sbreserve() already asserts that the lock is held, but it acts as implicit documentation and indicates that I looked at the function.
2017-07-18fix previous to make sxiccmu compile againJonathan Gray
2017-07-17The Linux kernel device trees use "allwinner,sun50i-h5-ccu" as the compatibleMark Kettenis
propert for the clock controller module. Prepare for the day when the u-boot device trees get update.
2017-07-17Implement a proper hash table instead of lumping evrything into a singleMark Kettenis
list.
2017-07-17Forbid overwriting a grant table entry currently in useMike Belopuhov
2017-07-17Reduce amount of CAS attempts in a busy-wait loop by a factor of 100Mike Belopuhov
2017-07-17Pick the correct TX buffer when unrollingMike Belopuhov
2017-07-17Reimplement mbuf/map to descriptor mappingMike Belopuhov
Previously descriptors have referenced DMA maps too loosely which led to bus_dmamap_unload being called before all fragments have been completed. This eliminates the last instance of excessive looping while waiting for a grant table entry to become available.
2017-07-17Make athn(4) send RTS frames at 2 Mbit/s again. I changed this to 1 Mbit/sStefan Sperling
in r1.38 of this file, however for some reason the hardware sends RTS frames at rates other than 1 Mbit/s which of course doesn't help. Not sure what is causing this problem, but the old behaviour worked better, so revert. ok krw@
2017-07-17Use .c instead of .o here too.Mark Kettenis
ok espie@
2017-07-17Grant table entries must be disassociated once CCB is completeMike Belopuhov
Attempting to unload the map before all chunks are done may lead to unnecessary looping.
2017-07-16Fix missing splx() in iwm_newstate_task(). Bug introduced 2 hours ago in r1.200.Stefan Sperling
2017-07-16Add WITNESS supportPhilip Guenther
ok visa@ kettenis@
2017-07-16Add comments to the definitions of iwm(4) driver-private flags.Stefan Sperling
2017-07-16Clear the framepointer (r3) in switch_trampoline to avoid faults in stack tracePhilip Guenther
Reduce size of switch stack frame to match normal 64byte alignment Eliminate useless 'osp' variable when setting up initial frame in cpu_fork() ok visa@ kettenis@
2017-07-16Remove the AP from iwm's firmware station table when moving out of AUTH state.Stefan Sperling
Fixes "could not remove MAC context" errors seen during AUTH->SCAN transitions.
2017-07-16Some net80211 state changes are triggered by incoming frames, and it isStefan Sperling
possible for such frames to move iwm(4) to the same state, e.g. AUTH->AUTH. Do not bother the firmware with such no-op state changes because that runs a high risk of hitting firmware errors. Problem reported by Gregor Best.
2017-07-16Match /clocks node on Allwinner H5.Mark Kettenis
2017-07-16Explain the multi-touch tracking function.Ulf Brosziewski
2017-07-16Minor changes in the setup.Ulf Brosziewski
1. Add middle-button areas to the clickpad defaults. 2. Handle the edge areas more uniformly. 3. Don't disable sw buttons at the bottom edge.
2017-07-16Pass the linker from make environment to makegap.sh. This is (only)Visa Hankala
a partial solution to cross compiling the kernel with KARL. OK deraadt@ a while ago
2017-07-16regenMark Kettenis
2017-07-16Update Skylake Graphics IDs.Mark Kettenis
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-24 01:51:01 +0000