| created | 2021-01-30T23:04:14Z |
|---|---|
| begin | 2021-01-25T00:00:00Z |
| end | 2021-01-26T00:00:00Z |
| path | src/sys |
| commits | 12 |
| date | 2021-01-25T01:45:55Z | |||
|---|---|---|---|---|
| author | dlg | |||
| files | src/sys/dev/pci/if_mcx.c | log | diff | annotate |
| message |
use an intrmap when establishing interrupts for queues. mcx is still hardcoded/limited to 1 queue for now, but this lets different mcx devices use different cpus for handling packets. looks good jmatthew@ |
|||
| date | 2021-01-25T03:40:46Z | |||
|---|---|---|---|---|
| author | dlg | |||
| files | src/sys/netinet/in_pcb.c | log | diff | annotate |
| src/sys/netinet/in_pcb.h | log | diff | annotate | |
| src/sys/netinet/tcp_output.c | log | diff | annotate | |
| message |
if stoeplitz is enabled, use it to provide a flowid for tcp packets. drivers that implement rss and multiple rings depend on the symmetric toeplitz code, and use it to generate a key that decides with rx ring a packet lands on. if the toeplitz code is enabled, this diff has the pcb and tcp layer use the toeplitz code to generate a flowid for packets they send, which in turn is used to pick a tx ring. because the nic and the stack use the same key, the tx and rx sides end up with the same hash/flowid. at the very least this means that the same rx and tx queue pair on a particular nic are used for both sides of the connection. as the stack becomes more parallel, it will also help keep both sides of the tcp connection processing in the one place. |
|||
| date | 2021-01-25T03:40:47Z | |||
|---|---|---|---|---|
| author | dlg | |||
| files | src/sys/netinet6/in6_pcb.c | log | diff | annotate |
| message |
if stoeplitz is enabled, use it to provide a flowid for tcp packets. drivers that implement rss and multiple rings depend on the symmetric toeplitz code, and use it to generate a key that decides with rx ring a packet lands on. if the toeplitz code is enabled, this diff has the pcb and tcp layer use the toeplitz code to generate a flowid for packets they send, which in turn is used to pick a tx ring. because the nic and the stack use the same key, the tx and rx sides end up with the same hash/flowid. at the very least this means that the same rx and tx queue pair on a particular nic are used for both sides of the connection. as the stack becomes more parallel, it will also help keep both sides of the tcp connection processing in the one place. |
|||
| date | 2021-01-25T09:11:36Z | |||
|---|---|---|---|---|
| author | yasuoka | |||
| files | src/sys/net/if_wg.c | log | diff | annotate |
| message |
Fix wg(4) ioctl to be able to handle multiple wgpeers. Diff from Yuichiro NAITO. ok procter |
|||
| date | 2021-01-25T09:36:48Z | |||
|---|---|---|---|---|
| author | dlg | |||
| files | src/sys/dev/pci/if_mcx.c | log | diff | annotate |
| message |
don't lose the M_FLOWID flag if the ipv4 cksum is ok. found while poking around with hrvoje popovski yes jmatthew@ |
|||
| date | 2021-01-25T11:11:22Z | |||
|---|---|---|---|---|
| author | dlg | |||
| files | src/sys/dev/pci/if_ixl.c | log | diff | annotate |
| message |
if the rx descriptor reports the rss hash, use it for the mbuf flowid. ok jmatthew@ |
|||
| date | 2021-01-25T12:27:42Z | |||
|---|---|---|---|---|
| author | dlg | |||
| files | src/sys/dev/pci/if_mcx.c | log | diff | annotate |
| message |
raise the max number of queues/interrupts to 16, up from 1. jmatthew@ has tried this before, but hrvoje popovski experienced breakage so it wasn't enabled. we've tightened the code up since then so it's time to try again. this diff has been tested by hrvoje popovski and myself ok jmatthew@ |
|||
| date | 2021-01-25T14:05:57Z | |||
|---|---|---|---|---|
| author | mglocker | |||
| files | src/sys/dev/usb/usbdi_util.c | log | diff | annotate |
| src/sys/dev/usb/usbdi_util.h | log | diff | annotate | |
| message |
Add the new function usbd_clear_endpoint_feature() which allows to issue an UR_CLEAR_FEATURE request on a specific endpoint address without the need to have a pipe open to that endpoint. From NetBSD, ok mpi@ |
|||
| date | 2021-01-25T14:14:42Z | |||
|---|---|---|---|---|
| author | mglocker | |||
| files | src/sys/dev/usb/ugen.c | log | diff | annotate |
| src/sys/dev/usb/uhidev.c | log | diff | annotate | |
| message |
Resolve data toggle out of sync problem for ugen(4) and uhidev(4) devices on xhci(4) controllers by clearing the interface endpoints before opening the pipes. Tested by Mikolaj Kucharski for ugen(4) and gnezdo@ for uhidev(4), plus myself for both. ok mpi@ |
|||
| date | 2021-01-25T19:37:17Z | |||
|---|---|---|---|---|
| author | kettenis | |||
| files | src/sys/arch/arm64/arm64/machdep.c | log | diff | annotate |
| src/sys/arch/arm64/arm64/pmap.c | log | diff | annotate | |
| message |
Give machdep.c a thorough cleanup that is long overdue. ok patrick@ |
|||
| date | 2021-01-25T19:37:18Z | |||
|---|---|---|---|---|
| author | kettenis | |||
| files | src/sys/arch/arm64/arm64/trap.c | log | diff | annotate |
| message |
Give machdep.c a thorough cleanup that is long overdue. ok patrick@ |
|||
| date | 2021-01-25T19:47:16Z | |||
|---|---|---|---|---|
| author | mvs | |||
| files | src/sys/net/bridgectl.c | log | diff | annotate |
| src/sys/net/bridgestp.c | log | diff | annotate | |
| src/sys/net/if_bridge.c | log | diff | annotate | |
| src/sys/net/if_bridge.h | log | diff | annotate | |
| message |
We have this sequence in bridge(4) ioctl(2) path: ifs = ifunit(req->ifbr_ifsname); if (ifs == NULL) { error = ENOENT; break; } if (ifs->if_bridgeidx != ifp->if_index) { error = ESRCH; break; } bif = bridge_getbif(ifs); This sequence repeats 8 times. Also we don't check value returned by bridge_getbig() before use. Newly introduced bridge_getbig() function replaces this sequence. This not only reduces duplicated code but also makes `bif' dereference safe. ok bluhm@ |
|||