Changes in usb

Changes in usb http://nxr.netbsd.org/rss/src/sys/dev/usb/ en Copyright 2005 Java Whitespace Thu Sep 10 05:00:00 UTC 2026 skrll Fix typo (s/untill/until/) in various comments. Thu Sep 03 02:00:00 UTC 2026 gutteridge don't expose us to userland until things are fully initialized. Same class of bugs that run(4) had in other USB drivers. PR kern/60665: run: hot-plugging USB wifi crashes due to use of if_attach Sat Aug 29 01:00:00 UTC 2026 maya run: don't expose us to userland until things are fully initialized. Fixes crash seen when hot-plugging run(4) with a running wpa_supplicant. PR kern/60665: run: hot-plugging USB wifi crashes due to use of if_attach Sat Aug 29 01:00:00 UTC 2026 maya motg: support control transfers (endpoint zero) aborts. Tested by Brook Milligan. Fri Jul 31 15:00:00 UTC 2026 skrll usmsc(4): Check the return value of smsc_readreg() in smsc_uno_read_reg() Fixes PR kern/58634: - Check the return value of smsc_readreg() in smsc_uno_read_reg() and return the error to the caller. - Return 0 as the register value when error occurred just for the consistency in smsc_uno_read_reg(). Note that it's not required because usbnet.c::uno_read_reg() doesn't access the value if error occurred. Mon Jul 06 11:00:00 UTC 2026 msaitoh xhci: Add missing UHF_PORT_POWER support to xhci_roothub_ctrl_locked() Wed Jun 17 09:00:00 UTC 2026 msaitoh fix various typos in comments. Mon Jun 01 20:00:00 UTC 2026 andvar ehci(4): work around AMD SB600 first control transfer failure Reload EHCI_ASYNCLISTADDR once after linking the first address-0 control qTD chain, for AMD SB600 quirk observed on my ASRock M3A-UCC. Addresses PR/57359. Sat May 02 07:00:00 UTC 2026 tsutsui if_ure.c: support pause frames with 8153* variants Sat Apr 18 00:00:00 UTC 2026 gutteridge sort UICLASS_* defines Mon Mar 23 09:00:00 UTC 2026 jmcneill regen Wed Mar 18 05:00:00 UTC 2026 mrg add korg as a vendor, two of their synths, and an amd root hub. Wed Mar 18 05:00:00 UTC 2026 mrg Allocate a cache line per ehci_sitd_t and drop the use of USBMALLOC_COHERENT in ehci_alloc_sitd to match the comment. Spotted by jmcneill@ Tue Mar 03 13:00:00 UTC 2026 skrll fix various typos in comments. Mon Feb 23 20:00:00 UTC 2026 andvar usbnet(9): Tighten tx path. 1. Verify, don't just assert, that the packet length is below the buffer size. Even if all the callers enforce the interface's MTU, I can't prove that the usbnet(9) tx buffer size is an upper bound enforced on the interface's MTU. We can remove the check later if we do enforce that upper bound at some point, which would probably be worth doing anyway since the MTU is checked earlier in the tx path. 2. Assert, don't check, that c->unc_xfer is nonnull. We can only reach the tx path if we cross if_init=usbnet_init_rx_tx, and that (via usbnet_tx_list_init) is guaranteed to fail and back out unless all of the usbnet_chain unc_xfers get initialized. 3. If we can't fit the packet into a buffer, drop it -- don't leave it in the queue to try again when it still won't fit in the buffer. 4. If the transfer for this packet fails, drop it -- don't leave it in the queue to try again just in case it might work better the next time. PR kern/59940: usbnet(9): uno_tx_prepare buffer overrun audit PR kern/59943: usbnet(9) keeps failed packet in queue to retry indefinitely Sun Feb 01 03:00:00 UTC 2026 riastradh ncm(4): Verify payload _and header_ fits in buffer. usbnet(9) only guarantees payload fits in buffer, doesn't know anything about the ncm(4) header overhead. PR kern/59940: usbnet(9): uno_tx_prepare buffer overrun audit Mon Jan 26 01:00:00 UTC 2026 riastradh Keep cur,next and queued values in sync, even when dying. Sat Jan 10 09:00:00 UTC 2026 mlelstv Also use atomic operation to store pipe pointers, not just to clear them. Probably fixes PR 56194. Sat Jan 10 08:00:00 UTC 2026 mlelstv urndis(4): Match more interface ids. There are some others we should consider too, based on what OpenBSD, FreeBSD, and Linux match, but I haven't tested with these devices: class subclass protocol 1. 0x02 (CDC) 0x02 (abstract control model) 0xff (? rndis?) 2. 0xef (misc) 0x01 (sync) 0x01 (active) 3. 0xef (misc) 0x04 (rndis) 0x03 (wimax) 4. 0xef (misc) 0x04 (rndis) 0x04 (wwan) Note: FreeBSD uses `UIPROTO_RNDIS' for (3) even though the USB-IF registry lists class 0xef, subclass 0x04, protocol 0x04 as `RNDIS over WiMAX'; likewise `UIPROTO_ACTIVESYNC' for (4) even though it's listed as `RNDIS over WWAN'. My guess is that for class 0xef subclass 0x04, _any_ protocol will really be RNDIS, and it was a mistake for FreeBSD to use those protocol numbers (which context suggests were supposed to apply to different subclasses). PR kern/59872: urndis(4): missing support for some devices Thu Jan 01 00:00:00 UTC 2026 riastradh