mmc: frequency too low to read EXT_CSD

2018-04-26 Thread Patrick Wildt
Hi, the SD/MMC supports two kinds of cards in memory mode as mass storage. MMCs and SD cards. In May 2016 kettenis@ changed the SD card code to bump the frequency to a higher level that is required to be supported by all SD cards, since on his controller the card didn't react if the frequency was

Re: Faster msdosfs read

2018-04-26 Thread Landry Breuil
On Thu, Apr 26, 2018 at 10:01:25PM +0200, Martin Pieuchot wrote: > This is the 4th attempt to implement clustering for msdosfs. See [0] > for the previous story. > > This version implements the strict necessary to read a maximum of 64k > per read. Unlike the previous version we no longer mix clu

tech@openbsd.org

2018-04-26 Thread Artturi Alm
On Thu, Apr 26, 2018 at 02:05:44PM +0200, Mark Kettenis wrote: > > Date: Thu, 26 Apr 2018 10:35:45 +0300 > > From: Artturi Alm > > > > Hi, > > > > wrote simple gpio indicator sensor driver, and spotted this. > > > > decided to keep the /* XXX */ for tech@, so this should be 'safe' > > for anyon

Re: options(4) improvements

2018-04-26 Thread Jason McIntyre
On Tue, Apr 24, 2018 at 07:13:01PM +0300, IL Ka wrote: > Hello Jason, > here is new patch. > > as indicated, now fixed. thanks, jmc > > diff --git share/man/man4/options.4 share/man/man4/options.4 > index 9ef493cf6..b0b0605c2 100644 > --- share/man/man4/options.4 > +++ share/man/man4/options

Faster msdosfs read

2018-04-26 Thread Martin Pieuchot
This is the 4th attempt to implement clustering for msdosfs. See [0] for the previous story. This version implements the strict necessary to read a maximum of 64k per read. Unlike the previous version we no longer mix cluster and block numbers. This speeds up msdosfs reads significantly. Regre

restore: preserve symlink mode

2018-04-26 Thread Philip Guenther
When a symlink is restored, set its mode to the value from the dump. This is totally cosmetic, as the kernel ignores the mode on symlinks, but it's nice to restore to the original. ok? Philip Index: tape.c === RCS file: /cvs/src

Re: Off-by-one in xhci

2018-04-26 Thread Paul Irofti
On Thu, Apr 26, 2018 at 05:44:18PM +0200, Martin Pieuchot wrote: > In xhci_xfer_tdsize() makes sure we do not forget to count packets with > a size inferior as UE_GET_SIZE(mps). This is only really used by the > isoc code, so the remaining calculation can now be corrected. > > While here use UE_G

Off-by-one in xhci

2018-04-26 Thread Martin Pieuchot
In xhci_xfer_tdsize() makes sure we do not forget to count packets with a size inferior as UE_GET_SIZE(mps). This is only really used by the isoc code, so the remaining calculation can now be corrected. While here use UE_GET_SIZE() coherently and howmany() where it applies. Finally remove a KASS

Re: Move FREF() inside fd_getfile()

2018-04-26 Thread Visa Hankala
On Thu, Apr 26, 2018 at 08:34:08AM +0200, Martin Pieuchot wrote: > On 25/04/18(Wed) 17:07, Visa Hankala wrote: > > On Wed, Apr 25, 2018 at 12:12:29PM +0200, Martin Pieuchot wrote: > > > The goal is to avoid races between fd_getfile() and FREF(). So we want > > > a properly refcounted 'struct file

tech@openbsd.org

2018-04-26 Thread Mark Kettenis
> Date: Thu, 26 Apr 2018 10:35:45 +0300 > From: Artturi Alm > > Hi, > > wrote simple gpio indicator sensor driver, and spotted this. > > decided to keep the /* XXX */ for tech@, so this should be 'safe' > for anyone not using gpioctl(8) no matter what :) > > i don't really see any risk, only w

Re: config -e vs ugen(4) fix

2018-04-26 Thread Alexandre Ratchov
On Tue, Apr 24, 2018 at 02:00:02PM +0200, Martin Pieuchot wrote: > Instead of having to disable USB drivers to force some devices to attach > as ugen(4), we can whitelist them. Diff below does that by adding a new > option to usbdevs(8): > > # usbdevs -l 7047:04ca -l a001:1199 > > The logic is

Re: net80211: stub SIOCS80211SCAN, make ifconfig scan instant

2018-04-26 Thread Klemens Nanni
On Wed, Apr 25, 2018 at 08:02:01PM +0200, Stefan Sperling wrote: > On Wed, Apr 25, 2018 at 08:55:26PM +0300, Paul Irofti wrote: > > Currently tested on iwm(4), urtwn(4), run(4) and athn(4). We are > > interested in wider testing, specially on iwn(4) and athn(4) in client > > mode. > > To clarify,

tech@openbsd.org

2018-04-26 Thread Artturi Alm
Hi, wrote simple gpio indicator sensor driver, and spotted this. decided to keep the /* XXX */ for tech@, so this should be 'safe' for anyone not using gpioctl(8) no matter what :) i don't really see any risk, only w/o one might need to do this w/ext resistors. comments? -Artturi diff --git