... List wrote: > >>> [SCSI] fix media change events for polled devices > >>> > >>> Commit: ... by > >> a341cd0f, by removing the event filter controlled by userspace. > > > >> Did ... pointing out that your suggested fix wouldn't work ... I think ...

linux-kernel - James Bottomley - Mar 21 2008 - 17:01

... List wrote: > > [SCSI] fix media change events for polled devices > > > > Commit: ... I just #if'd out the filter code to get polled ... s > needs, and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) > as obviously broken. James ...

linux-kernel - James Bottomley - Mar 21 2008 - 15:44

... List wrote: >>> [SCSI] fix media change events for polled devices >>> >>> Commit: ... is removed here, to make >>> events, we are depending on, happen ... by >> a341cd0f, by removing the event filter controlled by userspace. > >> Did ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 16:36

... Kernel Mailing List wrote: > [SCSI] fix media change events for polled devices > > Commit: > a341cd0f ... sr/sd's needs, and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) as obviously broken. Jeff -- To ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 13:12

... Linux Kernel Mailing List wrote: > > [SCSI] fix media change events for polled devices > > > > Commit: > > ... to sr/sd's > needs, and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) > as obviously broken. You mean adding a ...

linux-kernel - Kay Sievers - Mar 21 2008 - 14:33

... sd's needs, > > and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) as obviously > > broken. > > Patches, ... in with scsi-rc-fixes basically makes the whole lot work for ... .25 (both AN media change events and traditional polled ones). The ...

linux-kernel - James Bottomley - Mar 21 2008 - 15:38

... Kernel Mailing List wrote: >>> [SCSI] fix media change events for polled devices >>> >>> Commit: >>> ... to sr/sd's >> needs, and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) >> as obviously broken. > > You mean ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 16:42

... sd's needs, >>> and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) as obviously >>> broken. >> Patches, ... with scsi-rc-fixes basically makes the whole > lot work for 2 ... .25 (both AN media change events and traditional polled > ones). The ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 16:54

... Mailing List wrote: > >>> [SCSI] fix media change events for polled devices > >>> > >>> Commit: ... is removed here, to make > >>> events, we are depending on, happen ... pass the information on to drivers. For these, we set the attribute ...

linux-kernel - James Bottomley - Mar 21 2008 - 17:04

... List wrote: >>>>> [SCSI] fix media change events for polled devices >>>>> >>>>> Commit: ... is removed here, to make >>>>> events, we are depending on, happen ... was intended to be writable. > For starters, we have transport classes ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 17:10

... supported_events mask. > >> > >> And that is the fix -- initialize supported_events according to sr/sd' ... . Every device that detects a media change should send such event, so ... , regardless if the device is periodically polled, or supports AN. Thanks, ...

linux-kernel - Kay Sievers - Mar 21 2008 - 16:57

... . Yes, that's why we're having the argument ... it's ... Anyway, to recap... > > before your "fix": > > userspace interface always reflected list ... why we can't simply change it to 1 wholesale because ... 1 for AN) and let HAL take the polled events via ...

linux-kernel - James Bottomley - Mar 21 2008 - 18:52

... 's needs, >> and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) as obviously >> broken. > > Patches, ... +0200 [SCSI] sr,sd: send media state change modification events [...] [jejb: ... for new event API] so I'm hoping he will fix ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 16:39

... , 1 for AN) and let HAL > take the polled events via udev ... (or some other name) > file, fix the sysfs ro attribute and make ... was disabled rather than fixed, given that that change came first and arguably the follow-on change (285e9670) was an abuse of the ...

linux-kernel - Jeff Garzik - Mar 24 2008 - 23:09

... 1 for AN) and let HAL > > take the polled events via ... and isn't a regression. > > > > For 2.6.26 we can ... disabled rather than fixed, given that that change came > first and ... instead, that would be easier for me. Thanks, James -- To unsubscribe ...

linux-kernel - James Bottomley - Mar 24 2008 - 23:18

... to add a new file for it anyway. > > Previously, the events could be expected to ... to distinguish them ... but nothing's changed about the way it works. > Now ... is really moot ... we just need polling to work. James -- To unsubscribe from this ...

linux-kernel - James Bottomley - Mar 21 2008 - 17:49

... >> James Bottomley wrote: >>> It's fixed, and nothing else is additionally broken ... add a new file for it anyway. Previously, the events could be expected ... are essentially sending out-of-band events, because the receiver requested supported_events -- ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 17:13

... in the absence of AN. Anyway, to recap... before your "fix": userspace interface always reflected list of events sent via my new API after your "fix": events may or may not be reflected in userspace interface, ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 18:35

... Mar 2008, Jeff Garzik wrote: > > And that is the fix -- initialize supported_events according to sr/sd's needs, > and revert this change (4d1566ed2100d074ccc654e5cf2e44cdea3a01d0) as obviously > broken. Patches, please? I can do ...

linux-kernel - Linus Torvalds - Mar 21 2008 - 13:31

James Bottomley wrote: > It's fixed, and nothing else is additionally broken by the current fix. Other than the newly-inconsistent, exported-to-userspace interface, ITYM. Jeff -- To unsubscribe ...

linux-kernel - Jeff Garzik - Mar 21 2008 - 17:04