Re: hciconfig hci0 reset bug
Marcel Holtmann <[email protected]>
| Newsgroups | gmane.linux.bluez.user |
|---|---|
| Message-ID | <[email protected]> |
Hi Peter, >> So, if I'm reading this correctly, you should issue the reset for all >> CSR-based dongles with build id > 118. >> >> Am I on the right track here? Does this information look accurate >> to you >> guys? > > This is basically correct in that the fix appeared in firmware > around then > (the internal log says 117, but possibly that was never released--- > I've > confined my search to the development side). > > However, it's not that simple in that releases of earlier firmware > branches > were still being made for some time. Our ID numbers increase > monotonically > so these would have later numbers. I haven't done an exhaustive > search; > mostly these releases had L2CAP + RFCOMM on chip for embedded > applications, > so you wouldn't care about them. I don't see any evidence of a > later HCI > release based on the old code, offhand (but don't take this as > gospel). this is what Steven told me. We don't have to worry about the RFCOMM builds since they don't include a HCI interface. > If you want to be quite sure not to pick up a build which didn't > have the > fix for reset, you could restrict the change to BlueCore 2 and later > chips > using the "bcget chipver" call. The early branches without the fix > only ever ran on BlueCore 1: the fix was made in major release 12 and > BlueCore 2 first ran on major release 14. No vendor specific setup/init quirks that would have to call vendor commands :) Regards Marcel ------------------------------------------------------------------------- Check out the new SourceForge.net Marketplace. It's the best place to buy or sell services for just about anything Open Source. http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace