CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2020/02/13 08:39:02 Modified files: sys/dev/pv : xen.c Log message: Resolve a few issues with interrupt handling Upon a failed task_add, the adjusted reference counter needs to be decremented. xen_intr_schedule should follow the same logic as xen_intr. Tested by Niklas Hallqvist. Ok mpi@
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2019/12/13 03:29:05 Modified files: sys/dev/pv : Tag: OPENBSD_6_6 xen.c Log message: Revert: Always release interrupt source when unmasking the interrupt
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2019/12/12 19:16:53 Modified files: sys/dev/pv : xen.c Log message: Always release interrupt source when unmasking the interrupt xen_intr_unmask_release was not decrementing the reference counter on the interrupt source structure when bailing out early which led to the refcnt overflow. >From niklas, ok mlarkin
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2019/12/12 19:14:55 Modified files: sys/dev/pv : Tag: OPENBSD_6_6 xen.c Log message: Always release interrupt source when unmasking the interrupt xen_intr_unmask_release was not decrementing the reference counter on the interrupt source structure when bailing out early which led to the refcnt overflow. >From niklas, ok mlarkin
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2019/10/07 09:36:02 Modified files: sys/dev/pv : hyperv.c Log message: Attach Hyper-V guest services to VMBus 4.0 Tested by Andre Stoebe and Remi Locherer. ok deraadt, tb
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/06/18 13:05:24 Modified files: sys/dev/usb: uplcom.c Log message: Add an HP LD220 pole display Diff from Paul de Weerd (weerd@), tested by Jan-Piet Mens, thanks!
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/06/18 13:03:57 Modified files: sys/dev/usb: usbdevs.h usbdevs_data.h Log message: regen
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/06/18 13:03:33 Modified files: sys/dev/usb: usbdevs Log message: Add an HP LD220 pole display Diff from Paul de Weerd (weerd@), tested by Jan-Piet Mens, thanks!
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/06/14 12:44:43 Modified files: sys/arch/amd64/amd64: aesni.c Log message: Grab the FPU lock before calling aesni_enc ok guenther
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/05/17 06:32:33 Modified files: sys/dev/pv : if_hvn.c Log message: Let hvn_iff handle promisc mode activation Tested by Daniel Wade, , thanks!
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/05/17 02:09:26 Modified files: sys/arch/amd64/amd64: vmm.c Log message: Make VMM_DEBUG kernels compile again; ok mlarkin
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/05/15 03:21:52 Modified files: sys/net: ppp_tty.c Log message: Take NET_LOCK for pppioctl when called via the tty ioctl handler ok mpi
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/04/13 08:09:42 Modified files: sys/net: hfsc.c pf_ioctl.c Log message: Remove compatibility with pfctl from 6.1 and plug a few leaks No objections from henning, OK visa
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/04/08 12:26:29 Modified files: sys/arch/amd64/amd64: tsc.c Log message: Remove debugging printf ok deraadt, guenther
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/04/08 10:45:12 Modified files: sys/kern : kern_event.c Log message: Test for preexisting conditions when re-enabling events When an event that was disabled by EV_DISABLE or EV_DISPATCH is registered again, an associated filter must be ran to mark it active if a preexisting condition is present. The issue was reported and fix tested by Lukas Larsson , thanks! ok mpi
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/03/11 08:31:34 Modified files: sys/dev/pv : if_hvn.c Log message: Workaround an Azure UDP transmit checksum offloading bug According to https://reviews.freebsd.org/D12429, UDP transmit checksum offloading doesn't work in Azure when sizeof(IP hdr + UDP hdr + payload) is greater than 1420 and don't fragment bit isn't set in the IP header. Instead of peeking into the packet and attempting to fix the fallout, we restrict UDP checksum offloading to NDIS versions newer than the one used in Azure (6.30) in an attempt to keep the feature enabled. Unfortunately, there's no way to tell Azure apart from a standalone Hyper-V to make this check more precise.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/02/21 14:09:57 Modified files: sys/arch/amd64/amd64: via.c sys/arch/i386/i386: via.c Log message: Mark VIA padlock as capable of dealing with ESN There are no actual changes to the driver since the software crypto driver is called to handle authentication operations. This enabled padlock to be used when tunnels are setup with iked(8). Tested by and OK fcambus
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/02/20 08:02:13 Modified files: sys/arch/amd64/amd64: via.c sys/arch/i386/i386: via.c Log message: Convert key length from bits to bytes Reported by Renaud Allard, fix tested by Renaud (i386) and fcambus@ (amd64). OK visa, fcambus
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/01/21 11:54:46 Modified files: sys/dev/pv : xen.c Log message: Release interrupt source if the interrupt task is already scheduled Tested by Imre Oolberg, thanks!
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/01/21 11:51:19 Modified files: sys/dev/pv : xen.c Log message: Improve panic message
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/01/18 04:43:20 Modified files: sys/dev/pv : pvbus.c Log message: Fix an off-by-one uncovered by the recent free(9) change >From canacar@, thanks!
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2018/01/05 12:05:09 Modified files: sys/crypto : siphash.c Log message: Sync with the code in libc OK millert; original commit message by tedu@: memcpy from the right place. at this point, the used variable is not relevant. from Mark Karpilovskij. ok millert
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/12/09 07:00:21 Modified files: sys/dev/pv : if_xnf.c Log message: No need for the txb_used flag anymore
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/12/07 09:52:21 Modified files: sys/netinet: tcp_subr.c Log message: Initialize tcp_secret in tcp_init The initialization of a secret SHA256 context for generating TCP initial sequence numbers is moved out of tcp_set_iss_tsm used to set up ISN for new connections and into tcp_init, sparing the need for a global flag. OK deraadt, visa, mpi
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/11/26 10:06:46 Modified files: sys/dev: rnd.c Log message: Don't mention XOR as a mix-in function since addition is done since 1.180
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/11/26 09:11:45 Modified files: sys/dev/pv : xen.c Log message: Xen interrupt task barrier is just a taskq_barrier so call it instead of rolling its own now that it's been made available. OK dlg
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/11/19 06:43:06 Modified files: sys/dev: rnd.c Log message: Remove interlocks between producers and consumers of randomness data A lot of randomness event producers are executed in the interrupt context increasing the time spent in the interrupt handler resulting in extra costs when adding randomness data to the pool. However, in practice randomness event producers require interlocking between each other, but not with with consumers due to the opportunistic nature of event consumers. To be able to take advantage of this idea, the ring buffer indexing is now done with two free running producer and consumer counters modulo power of 2 size of the ring buffer. With input from and OK visa, tb, jasper
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/11/07 17:37:18 Modified files: sys/dev/pv : if_xnf.c Log message: Fixup what looks like a merge mistake; no functional change
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/28 09:25:20 Modified files: usr.bin/netstat: mbuf.c Log message: Bring "netstat -m" output up to speed with the kernel The kernel no longer sets watermarks on cluster pools rendering "max" values useless. Instead, there's now a global limit on how much memory all cluster pools combined together can allocate from the system. The limit is set to kern.maxclusters number of 2Kb increments which allows us to display current, peak and maximum total memory used by the network. OK claudio, millert
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/25 16:29:41 Modified files: sys/arch/amd64/amd64: fpu.c Log message: Initialize processor extended state in fpu_kernel_enter Tested by a few; OK visa, sthen
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/24 08:49:29 Modified files: sys/netinet: tcp_input.c tcp_var.h Log message: Refactor handling of partial TCP acknowledgements With input from Klemens Nanni, OK visa, mpi, bluhm
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/23 09:41:29 Modified files: sys/arch/amd64/amd64: machdep.c sys/arch/amd64/include: cpu.h Log message: Expose TSC properties via sysctl nodes >From Adam Steen with suggestions from tedu@ OK deraadt, mlarkin (as part of a larger diff)
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/22 08:11:34 Modified files: sys/conf : GENERIC sys/netinet: tcp_input.c tcp_output.c tcp_subr.c tcp_timer.c tcp_usrreq.c tcp_var.h Log message: Unconditionally enable TCP selective acknowledgements (SACK) OK deraadt, mpi, visa, job
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/19 16:09:49 Modified files: sys/arch/amd64/amd64: tsc.c Log message: Don't forget to set the timecounter frequency Found the hard way, fix tested and OK mlarkin
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/18 09:26:05 Modified files: sys/arch/amd64/amd64: tsc.c Log message: Fixup previous; noticed by canacar@
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/18 06:52:06 Modified files: sys/arch/amd64/amd64: identcpu.c tsc.c Log message: Set TSC timecounter frequency to the CPU frequency estimate if unknown ok mlarkin
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/17 07:58:10 Modified files: sys/arch/amd64/amd64: tsc.c Log message: Correctly handle frequency measurement failures Issue observed by Joe Gidi and canacar@. While here, make sure that the success rate is at least 2 out of 3 measurements but with an increased threshold.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/16 09:09:43 Modified files: lib/libc/gen : sysctl.3 Log message: draft-ietf-tcpm-initcwnd was published as rfc 6928
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/13 17:41:34 Modified files: share/man/man5 : pf.conf.5 Log message: Integrate the description of flow queues into the main body of text
Re: CVS: cvs.openbsd.org: src
On Fri, Oct 06, 2017 at 13:33 +, Mike Belopuhov wrote: > CVSROOT: /cvs > Module name: src > Changes by: mi...@cvs.openbsd.org 2017/10/06 07:33:53 > > Modified files: > sys/arch/amd64/amd64: identcpu.c > sys/arch/amd64/conf: files.amd64 > sys/arch/amd64/include: cpu.h cpuvar.h > sys/dev/acpi : acpihpet.c acpitimer.c > Added files: > sys/arch/amd64/amd64: tsc.c > > Log message: > Recalibrate TSC timecounter with HPET and PM timer > > If frequency of an invariant (non-stop) time stamp counter is measured > using an independent working timecounter that has a known frequency, we > can assume that the measured TSC frequency is as good as the resolution > of the timecounter that we use to perform the measurement. This lets us > switch from this high quality but expensive source to the cheaper TSC > without sacrificing precision on a wide range of modern CPUs. > > From Adam Steen <a...@adamsteen.com.au> with tweaks from reyk@ and myself. > > Tested by brynet@, sthen@ and others, OK mlarkin, sthen > ACHTUNG! Please make sure to do a "make config" when building the kernel to update offset tables because of the cpu_info structure changes.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/10/06 07:33:53 Modified files: sys/arch/amd64/amd64: identcpu.c sys/arch/amd64/conf: files.amd64 sys/arch/amd64/include: cpu.h cpuvar.h sys/dev/acpi : acpihpet.c acpitimer.c Added files: sys/arch/amd64/amd64: tsc.c Log message: Recalibrate TSC timecounter with HPET and PM timer If frequency of an invariant (non-stop) time stamp counter is measured using an independent working timecounter that has a known frequency, we can assume that the measured TSC frequency is as good as the resolution of the timecounter that we use to perform the measurement. This lets us switch from this high quality but expensive source to the cheaper TSC without sacrificing precision on a wide range of modern CPUs. >From Adam Steenwith tweaks from reyk@ and myself. Tested by brynet@, sthen@ and others, OK mlarkin, sthen
CVS: cvs.openbsd.org: www
CVSROOT:/cvs Module name:www Changes by: mi...@cvs.openbsd.org 2017/10/01 12:51:04 Modified files: . : 62.html Log message: Mention Xen improvements and FQ-CoDel support
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/09/26 16:12:04 Modified files: sys/dev/ic : wdc.c Log message: Prevent null pointer dereference when probing channels Account for the case when wdc is attached to the ISA bus and performs channel probing using a dummy structure that lacks the back pointer to the controller's softc. Bug reported and fix tested by Andrew Daugherity, thanks! OK phessler, jsg, krw, deraadt
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/09/18 13:45:52 Modified files: share/man/man4 : hyperv.4 Log message: Mention that Generation 2 Virtual Machines are currently not supported
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/23 04:10:56 Modified files: sys/dev/pci: if_txp.c Log message: Prevent trashing of the producer index and descriptor counter The last change has removed a clause that required caching producer index and descriptor counter but introduced an early break that can cause uninitialized values to be stored in the ring header. Coverity CID 1453170, 1452971 OK visa
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/17 12:22:43 Modified files: sys/net: hfsc.c Log message: Remove a class from parent's active list when queue is empty When hfsc_deq_begin can't obtain an mbuf from the queue which can happen with fq_codel, the hfsc_deq_commit isn't called and HFSC doesn't get to update the active queue list of the parent class. To successfully purge and destroy such class, we need to make sure to run the service curve update routine even if we haven't dequeued anything so that the parent class doesn't accumulate stale child classes on its active list. Reported by Atanas Vladimirov, OK visa
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/17 12:14:39 Modified files: sys/net: hfsc.c Log message: Resolve races with the hfsc_deferred timeout Since NET_LOCK no longer protects the ioctl path from timeouts running while ifq configuration is happening, the hfsc_deferred timeout needs to make sure that the underlying HFSC object is locked before attempting to reference it. This prevents the race happening after we release the ifq mutex in ifq_attach but before hfsc_free disables the timeout. The second race is possible after hfsc_alloc but before the ifq mutex gets locked in ifq_attach and a new set of ifq operations is installed. If hfsc_deferred fires in between, it wouldn't be able to reschedule itself and will no longer be executed. To prevent this, the timeout can be established when packets are enqueued instead of the moment the queue is created. With input and OK visa
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/17 04:14:08 Modified files: sys/net: if_bridge.c Log message: Skip SPD lookups for short packets on IPsec-enabled bridge When short packets are sent to the bridge with IPsec enabled, an incorrect error path can be taken which leads to a lookup of an SPD entry using an uninitialized SPI. Most of the time this will fail, however there's a chance that an existing SPD entry corresponds to the provided SPI which leads to use of another uninitialized variable used to offset the IP or IPv6 header in order to get to the security protocol header. ESP performs packet length checks and will fail when such packets will reach it, but AH and IPComp don't have similar checks and are affected the most. CID 1452946, 1452957; Severity: Major OK millert, visa, bluhm
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/16 11:52:17 Modified files: sys/kern : kern_bufq.c Log message: Correct the check when selecting an elevator Coverity CID 1453358; Severity: unlikely, not user-visible ok millert, visa
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/16 08:19:57 Modified files: sys/net: pf_table.c Log message: Validate pfra_type after copyin before using it to index an array Don't trust the value of pfra_type blindly since it's coming from userland and sanitize it in pfr_validate_addr that is called after every copyin and also perform the check in pfr_create_kentry before we attempt to use the value not after. Coverity CID 1452909, 1453097, 1453384; Severity: Minor It can be triggered only by root by default or anyone with write access to /dev/pf if such access is provided. ok visa, bcook, sashan, jsg
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/10 14:13:57 Modified files: sys/dev/pv : xen.c Log message: Don't forget to call va_end in xen_hypercall Coverity CID 1453343
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/10 12:14:56 Modified files: sys/dev/pv : xenstore.c Log message: Prevent an unlikely resource leak Coverity CID 1453069; Severity: unlikely, not user-visible.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/10 11:22:48 Modified files: sys/dev/pv : hvs.c Log message: Correct the upper bound for the command size before the passthrough Coverity CID 1453317; Severity: unlikely, not user-visible.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/10 11:11:49 Modified files: sys/dev/pv : hvs.c Log message: Buffer size for the SCSI vendor string should be 8 char long A vendor and part of the product string got copied into a larger on-stack buffer as a result of an out-of-bounds access, however only 4 characters are meaningful in this context. Coverity CID 1453206; Severity: insignificant.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/10 10:48:47 Modified files: sys/dev/pv : if_hvn.c Log message: Fixup upper bound for the completion descriptor identifier Coverity CID 1452864; Severity: unlikely, not user-visible.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/10 09:25:57 Modified files: sys/dev/pv : hypervic.c Log message: Fail if unknown type of an address family was specified Coverity CID 1452981; Severity: unlikely, not user-visible.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/08/10 09:01:42 Modified files: sys/dev/pci: mpii.c Log message: Remove a leftover bit that was dereferencing an uninitialized pointer Coverity CID 1453398; ok deraadt.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/24 09:20:46 Modified files: sys/net: hfsc.c Log message: Don't attempt to deactivate a class that doesn't have any packets in it Classes are considered "active" when they have packets assigned to them. Active classes are tracked on various lists and during the purge operation must be dissociated from them. During factoring out of the HFSC internal FIFO operations, the check for whether the class is empty got lost. Issue reported and fix tested by abieber@ and myself.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/23 19:23:46 Modified files: sys/net: fq_codel.c Log message: Fixup free list handling in fqcodel_deq_begin We're growing an mbuf free list while iterating over flow queues and need to adjust our internal statistics on every iteration by using a portion of the free list corresponding to the current iteration.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/21 14:58:07 Modified files: usr.sbin/hostctl: hostctl.8 Log message: Stress the fact that individual pvbus device nodes correspond to a single key-value store and if there are other ones available, different nodes must be used to access them.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/21 14:00:47 Modified files: sys/dev/pv : xen.c xenvar.h Log message: Replace MD _bus_dmamap_* function calls with MI ones
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/21 13:15:43 Modified files: etc/etc.amd64 : MAKEDEV etc/etc.i386 : MAKEDEV Log message: sync
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/21 13:15:07 Modified files: etc: MAKEDEV.common Log message: Multiple virtualization layers may be available at the same time reachable through different pvbus device nodes. Suggestion and OK deraadt, OK reyk
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/21 12:27:32 Modified files: share/man/man4 : xen.4 Log message: Document settings for XenServer Figured out by Dinar Talypov and Maxim Khitrov. With input from deraadt@ and schwarze@. OK schwarze.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/21 06:28:15 Modified files: share/man/man4 : ciss.4 Log message: Fixup too much indentation
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/20 15:14:39 Modified files: regress/sbin/pfctl: pf109.ok pf110.in pf110.ok Log message: Fixup pf109.ok and rewrite pf110
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/19 10:48:22 Modified files: sys/dev/pv : xen.c Log message: Turn this into a panic since there's no way to recover from it
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/19 10:31:56 Modified files: sys/dev/usb: if_urndis.c Log message: Add a handler for RNDIS status messages 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
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/19 07:41:20 Modified files: sys/net: hfsc.c Log message: Fix the test condition weakened a recent change
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/19 07:02:46 Modified files: regress/sbin/pfctl: Makefile Added files: regress/sbin/pfctl: pf105.in pf106.in pf107.in pf108.in pf109.in pf105.ok pf106.ok pf107.ok pf108.ok pf109.ok pf110.in pf111.in pf110.ok pf111.ok pfail58.in pfail59.in pfail58.ok pfail59.ok pfail60.in pfail61.in pfail62.in pfail60.ok pfail61.ok pfail62.ok Log message: Regression tests asserting various HFSC and FQ-CoDel functionality
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/19 06:58:31 Modified files: sbin/pfctl : parse.y pfctl.c pfctl_parser.c usr.bin/systat : pftop.c Log message: Allow HFSC classes to use flow queues The FQ-CoDel related configuration (flows, quantum) becomes available for the regular bandwidth queue. Internally the kernel will pick the FQ-CoDel for use as a queue manager for the specified class instead of the FIFO. Discussed with and OK henning@ at d2k17 as a part of a larger diff.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/19 06:54:09 Modified files: sys/net: hfsc.c Log message: Factor out internal FIFO queue manager 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.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/19 06:51:31 Modified files: sbin/pfctl : parse.y pfctl.c pfctl_queue.c sys/net: pf_ioctl.c pfvar.h Log message: Rework HFSC vs FQ-CoDel checks 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.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/18 05:42:51 Modified files: sys/dev/pci: xspd.c Log message: Fixup the printf string, spacing
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/17 10:32:26 Modified files: sys/dev/pv : xen.c Log message: Forbid overwriting a grant table entry currently in use
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/17 10:04:31 Modified files: sys/dev/pv : xen.c Log message: Reduce amount of CAS attempts in a busy-wait loop by a factor of 100
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/17 10:01:24 Modified files: sys/dev/pv : if_xnf.c Log message: Pick the correct TX buffer when unrolling
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/17 09:58:22 Modified files: sys/dev/pv : if_xnf.c Log message: Reimplement mbuf/map to descriptor mapping 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.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/17 04:30:03 Modified files: sys/dev/pv : xbf.c Log message: Grant table entries must be disassociated once CCB is complete Attempting to unload the map before all chunks are done may lead to unnecessary looping.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/14 15:53:34 Modified files: sys/dev/pv : xen.c Log message: Spacing
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/14 14:08:46 Modified files: sys/dev/pv : xen.c Log message: Reduce the number of CAS loops from ludicrous to ridiculous Now that the source of the delay with releasing grant table entries has been identified and fixed the number of attempts to CAS entry flags can be substantially reduced and while it's decreased by a factor of 10, it should go down at least a 100 more in the future.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/14 13:09:52 Modified files: sys/dev/pv : xen.c xenvar.h Log message: Silence the interrupt source until the interrupt task has done its job This small change significantly improves performance under load and halves the number of received interrupts.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/14 12:24:51 Modified files: sys/dev/pv : if_xnf.c Log message: Support out-of-order TX completion notifications Apparently, just like the disk interface, the network backend may post TX completions out-of-order, i.e. the ring position doesn't determine which descriptor is being completed and the code must look at the response id inside the completion message. It might seem obvious, but since networking hardware doesn't usually work this way, it was something that has been overlooked. Based on instability reports from Kirill Miazine , thanks!
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/13 05:23:53 Modified files: sys/dev/pv : hvs.c Log message: Add a comment explaining why pciide(4) must skip attaching disks Requested by kettenis@.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/12 08:36:57 Modified files: sys/dev/pci: if_txp.c if_txpreg.h Log message: Tweak txp to avoid ifq_deq_begin/commit/rollback Setup the transmit ring the way FreeBSD driver does giving each descriptor a 16 fragment chain and revamp the start routine to use m_defrag for long chains and avoid the dequeue begin/commit/ rollback dance. OK dlg
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/12 08:25:36 Modified files: sys/dev/pci: if_vic.c Log message: Reshuffle vic_start and get rid of the dequeue begin/rollback/commit dance Tested on ESXi 5.5.0, OK reyk
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/12 07:44:54 Modified files: sys/dev/pv : hvs.c Log message: Prefer paravirtualized interface for IDE disks as well
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/12 07:40:59 Modified files: sys/dev/ic : wdc.c wdcvar.h sys/dev/pci: pciide.c Log message: Introduce two quirks to prevent attaching ATA and ATAPI devices Hyper-V and Xen have varying support for detaching emulated IDE devices ranging from none on Hyper-V to only IDE disks but not CDROM (Xen). The quirk mechanism provides a way of enforcing the desired behavior. With suggestions and OK kettenis, tedu, mlarkin
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/07/12 05:13:22 Modified files: sys/kern : vfs_bio.c Log message: Invalidate read-ahead buffers when read short Buffercache performs read-ahead for cluster reads by extending the length of an original read operation to the MAXPHYS (64k). Upon I/O completion, the length is trimmed and the buffer is returned to the filesystem and the remaining data is cached. However, under certain circumstances, the underlying hardware may fail to do a complete I/O operation and return with a non- zero value of the residual length (i.e. data that wasn't read). The residual length may exceed the size of an original request and must be re-adjusted to uphold the contract with the caller, e.g. the filesystem. At the same time, read-ahead buffers that cover chunks of memory corresponding to the residual length must be invalidated and not cached. Discussed at length during d2k17, ok tedu
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/28 13:30:25 Modified files: sys/net: pf_ioctl.c pfvar.h Log message: Introduce a simple mechanism to select the appropriate queue manager Discussed with and OK henning@ at d2k17 as a part of a larger diff.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/28 13:27:42 Modified files: sys/net: pf_ioctl.c Log message: Tighten up FQ-CoDel vs HFSC checks PFQS_FLOWQUEUE is about to become a flag that HFSC classes may specify as well; thus a better way of telling FQ-CoDel and HFSC apart needs to be found. At the moment its derived from the queue specification. Discussed with and OK henning@ at d2k17 as a part of a larger diff.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/28 13:13:35 Modified files: sys/net: fq_codel.c Log message: Provide pf queue manager API Discussed with and OK henning@ at d2k17 as a part of a larger diff.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/28 12:31:03 Modified files: sys/net: pfvar.h Log message: Extend pf queueing ops to include queue manager hooks Discussed with and OK henning@ at d2k17 as a part of a larger diff.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/28 12:24:02 Modified files: sys/conf : files sys/net: hfsc.c Log message: hfsc.c should depend on pf instead of inet Discussed with and OK henning@ at d2k17 as a part of a larger diff.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/27 07:56:15 Modified files: sys/dev/pv : hyperv.c Log message: Provide a handler stub for the "channel rescind" message
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/26 14:12:14 Modified files: sys/dev/pv : hvs.c Log message: Minor cleanup; when polling delay before calling the interrupt handler
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/26 12:42:23 Modified files: sys/dev/pv : hyperv.c Log message: Rework the deferred interrupt loop By performing a task_add an interrupt handler can rely on the taskq_thread to invoke it again with an additional benefit of being able to sched_pause when required. In the long run more than 99.8% of calls do not require an additional iteration.
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/23 13:09:21 Modified files: sys/dev/pv : hyperv.c Log message: Event interrupt handler should call hv_channel_schedule now
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/23 13:07:13 Modified files: sys/dev/pv : hvs.c Log message: Request deferred interrupt processing
CVS: cvs.openbsd.org: src
CVSROOT:/cvs Module name:src Changes by: mi...@cvs.openbsd.org 2017/06/23 13:05:42 Modified files: sys/dev/pv : hyperv.c hypervvar.h Log message: Introduce deferred interrupt processing capability Hyper-V interrupts seem to be sometimes delivered before the message becomes available on the channel ring. This is reproducible on hvs(4) under load. This change is modelled on the workaround found in the Linux driver.