[opensuse-kernel] Tumbleweed kernel changes to help upstream development
Hi all, As part of one of the Linux kernel Summit discussions today, it was brought up that after a kernel is released (for example 3.5), it's a bit too late to be doing testing to see how well it is working out. The .0 release is usually a bit rough, and it takes until the .1 or .2 release to get most major issues out of the way. So, the kernel developers would like to get a wider range of testing, and one thing proposed would be to have rolling distros switch their kernel over a bit earlier to the new release than they had been doing. Specifically, around the -rc5 point in time would be great. That way, any reported regressions could be fixed sooner and get into the final .0 release for everyone to use. Now this does place a bit of a larger burden on the users of those distros to be diligent in reporting problems, and the distro engineers to report the issues upstream as well, but it sounds like a reasonable thing to try out. So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening? thanks, greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On 28 August 2012 06:05, Greg KH <gregkh@linux.com> wrote:
Hi all,
As part of one of the Linux kernel Summit discussions today, it was brought up that after a kernel is released (for example 3.5), it's a bit too late to be doing testing to see how well it is working out. The .0 release is usually a bit rough, and it takes until the .1 or .2 release to get most major issues out of the way.
So, the kernel developers would like to get a wider range of testing, and one thing proposed would be to have rolling distros switch their kernel over a bit earlier to the new release than they had been doing. Specifically, around the -rc5 point in time would be great. That way, any reported regressions could be fixed sooner and get into the final .0 release for everyone to use.
Now this does place a bit of a larger burden on the users of those distros to be diligent in reporting problems, and the distro engineers to report the issues upstream as well, but it sounds like a reasonable thing to try out.
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
I don't see any issue with switching earlier, other than finding some bugs sooner :-) Most people that run Tumbleweed are generally living on the edge of their seats so should be more than capable of reporting bugs. I find you proposal sane (well as sane as anything is), and as such should be implemented. Regards, Andy -- Andrew Wafaa IRC: FunkyPenguin GPG: 0x3A36312F -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Mon, Aug 27, 2012 at 10:15 PM, Andrew Wafaa <awafaa@opensuse.org> wrote:
On 28 August 2012 06:05, Greg KH <gregkh@linux.com> wrote:
Hi all,
As part of one of the Linux kernel Summit discussions today, it was brought up that after a kernel is released (for example 3.5), it's a bit too late to be doing testing to see how well it is working out. The .0 release is usually a bit rough, and it takes until the .1 or .2 release to get most major issues out of the way.
So, the kernel developers would like to get a wider range of testing, and one thing proposed would be to have rolling distros switch their kernel over a bit earlier to the new release than they had been doing. Specifically, around the -rc5 point in time would be great. That way, any reported regressions could be fixed sooner and get into the final .0 release for everyone to use.
Now this does place a bit of a larger burden on the users of those distros to be diligent in reporting problems, and the distro engineers to report the issues upstream as well, but it sounds like a reasonable thing to try out.
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
I don't see any issue with switching earlier, other than finding some bugs sooner :-) Most people that run Tumbleweed are generally living on the edge of their seats so should be more than capable of reporting bugs. I find you proposal sane (well as sane as anything is), and as such should be implemented.
Regards,
Andy
-- Andrew Wafaa IRC: FunkyPenguin GPG: 0x3A36312F -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
As a long-time "edge of the seats" Linux desktop/laptop user (I started at Red Hat 6.2 and ran Gentoo testing for years) I'm going to disagree. A rolling release is fine for anything *but* the kernel. As far as I'm concerned, the boot loader, kernel, filesystem checks, X windows, networking / wifi, sound cards, etc. - everything that does its magic up to the point where the display manager asks me to log in - needs to be stable and signed off on by some kind of human / machine QA process. I'd like to see some kernel software metrics - bug finding rates / reporting rates / closing rates, average number of days between the last kernel release candidate and stable, etc. I'm fine with GNOME, KDE, Firefox, LibreOffice rolling releases. Hell, I run most of my scientific applications and eBook publishing tools compiled from upstream tarballs and some things even from version control repositories! But the kernel - no! Once I have filesystems, network, audio and video working I don't want to be part of the troubleshooting efforts on them again. -- Twitter: http://twitter.com/znmeb; Computational Journalism Publishers Workbench: http://j.mp/QCsXOr How the Hell can the lion sleep with all those people singing "A weem oh way!" at the top of their lungs? -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
In SUSE Studio terms, if I build a Tumbleweed appliance with the "Server" template, I want everything in it solid. And I'd add the tool chain - gcc / make / bash / vi / grep and friends - to the list of things that need to be solid / non-beta whether or not they're in Tumbleweed. On Tue, Aug 28, 2012 at 9:40 AM, M. Edward (Ed) Borasky <znmeb@znmeb.net> wrote:
On Mon, Aug 27, 2012 at 10:15 PM, Andrew Wafaa <awafaa@opensuse.org> wrote:
On 28 August 2012 06:05, Greg KH <gregkh@linux.com> wrote:
Hi all,
As part of one of the Linux kernel Summit discussions today, it was brought up that after a kernel is released (for example 3.5), it's a bit too late to be doing testing to see how well it is working out. The .0 release is usually a bit rough, and it takes until the .1 or .2 release to get most major issues out of the way.
So, the kernel developers would like to get a wider range of testing, and one thing proposed would be to have rolling distros switch their kernel over a bit earlier to the new release than they had been doing. Specifically, around the -rc5 point in time would be great. That way, any reported regressions could be fixed sooner and get into the final .0 release for everyone to use.
Now this does place a bit of a larger burden on the users of those distros to be diligent in reporting problems, and the distro engineers to report the issues upstream as well, but it sounds like a reasonable thing to try out.
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
I don't see any issue with switching earlier, other than finding some bugs sooner :-) Most people that run Tumbleweed are generally living on the edge of their seats so should be more than capable of reporting bugs. I find you proposal sane (well as sane as anything is), and as such should be implemented.
Regards,
Andy
-- Andrew Wafaa IRC: FunkyPenguin GPG: 0x3A36312F -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org
As a long-time "edge of the seats" Linux desktop/laptop user (I started at Red Hat 6.2 and ran Gentoo testing for years) I'm going to disagree. A rolling release is fine for anything *but* the kernel. As far as I'm concerned, the boot loader, kernel, filesystem checks, X windows, networking / wifi, sound cards, etc. - everything that does its magic up to the point where the display manager asks me to log in - needs to be stable and signed off on by some kind of human / machine QA process.
I'd like to see some kernel software metrics - bug finding rates / reporting rates / closing rates, average number of days between the last kernel release candidate and stable, etc. I'm fine with GNOME, KDE, Firefox, LibreOffice rolling releases. Hell, I run most of my scientific applications and eBook publishing tools compiled from upstream tarballs and some things even from version control repositories! But the kernel - no! Once I have filesystems, network, audio and video working I don't want to be part of the troubleshooting efforts on them again.
-- Twitter: http://twitter.com/znmeb; Computational Journalism Publishers Workbench: http://j.mp/QCsXOr
How the Hell can the lion sleep with all those people singing "A weem oh way!" at the top of their lungs?
-- Twitter: http://twitter.com/znmeb; Computational Journalism Publishers Workbench: http://j.mp/QCsXOr How the Hell can the lion sleep with all those people singing "A weem oh way!" at the top of their lungs? -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Mon, 2012-08-27 at 22:05 -0700, Greg KH wrote:
Hi all,
As part of one of the Linux kernel Summit discussions today, it was brought up that after a kernel is released (for example 3.5), it's a bit too late to be doing testing to see how well it is working out. The .0 release is usually a bit rough, and it takes until the .1 or .2 release to get most major issues out of the way.
So, the kernel developers would like to get a wider range of testing, and one thing proposed would be to have rolling distros switch their kernel over a bit earlier to the new release than they had been doing. Specifically, around the -rc5 point in time would be great. That way, any reported regressions could be fixed sooner and get into the final .0 release for everyone to use.
Now this does place a bit of a larger burden on the users of those distros to be diligent in reporting problems, and the distro engineers to report the issues upstream as well, but it sounds like a reasonable thing to try out.
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
No strong objection, but a minor comment.. In practice, rc5 vs .0 probably won't make much difference at all, but more users will be bitten, else the move was a failure. So this move explicitly plans on users being bitten, detracting from Tumbleweed's rolling ~stable allure, _evading_ developmental regression detection and reporting duties being a major attractor to running rolling ~stable ;-) -Mike -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Tue, Aug 28, 2012 at 08:55:56AM +0200, Mike Galbraith wrote:
In practice, rc5 vs .0 probably won't make much difference at all, but more users will be bitten, else the move was a failure. So this move explicitly plans on users being bitten, detracting from Tumbleweed's rolling ~stable allure, _evading_ developmental regression detection and reporting duties being a major attractor to running rolling ~stable ;-)
That's a really good point. I guess I'm abusing my role here as both a kernel developer who wants to see stable kernels released by the community, and as the Tumbleweed maintainer, sorry about that. I'm all for helping kernel.org developers out, but at the expense of people who are counting on Tumbleweed to not mess their systems up, I probably shouldn't do that, as you and Lars and others point out. Let me talk to James Bottomley about this a bit more, he's the one that proposed doing this yesterday in the meeting, as well as talking with David from Fedora, and see if there's some other way we can come up with that can help out. Maybe just have a Tumbleweed:Kernel repo that people can add to get this newer kernel if they want to be on even more of a bleading edge, that would like to the proper Kernel:HEAD release at the -rc5 time period. Perhaps that might balance the needs of users better? greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Tue, 2012-08-28 at 06:46 -0700, Greg KH wrote:
On Tue, Aug 28, 2012 at 08:55:56AM +0200, Mike Galbraith wrote:
In practice, rc5 vs .0 probably won't make much difference at all, but more users will be bitten, else the move was a failure. So this move explicitly plans on users being bitten, detracting from Tumbleweed's rolling ~stable allure, _evading_ developmental regression detection and reporting duties being a major attractor to running rolling ~stable ;-)
That's a really good point. I guess I'm abusing my role here as both a kernel developer who wants to see stable kernels released by the community, and as the Tumbleweed maintainer, sorry about that.
It's a worthy goal, just a question (as others have mentioned) of what Tumbleweed's primary mission is. Defining _how_ close to the edge a rolling ~stable distribution should get is rather a sticky wicket.
I'm all for helping kernel.org developers out, but at the expense of people who are counting on Tumbleweed to not mess their systems up, I probably shouldn't do that, as you and Lars and others point out.
Let me talk to James Bottomley about this a bit more, he's the one that proposed doing this yesterday in the meeting, as well as talking with David from Fedora, and see if there's some other way we can come up with that can help out.
Maybe just have a Tumbleweed:Kernel repo that people can add to get this newer kernel if they want to be on even more of a bleading edge, that would like to the proper Kernel:HEAD release at the -rc5 time period. Perhaps that might balance the needs of users better?
Making kernel packages available in Tumbleweed and other stable bases sounds like a fine idea to me. That may be attractive to those who find bleeding edge everything (factory) to be far too much of an adventure. Recovering from a dud kernel is (generally) a walk in the park compared to userspace borkage. -Mike -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On 2012-08-28T06:46:46, Greg KH <gregkh@linux.com> wrote:
On Tue, Aug 28, 2012 at 08:55:56AM +0200, Mike Galbraith wrote:
In practice, rc5 vs .0 probably won't make much difference at all, but more users will be bitten, else the move was a failure. So this move explicitly plans on users being bitten, detracting from Tumbleweed's rolling ~stable allure, _evading_ developmental regression detection and reporting duties being a major attractor to running rolling ~stable ;-)
That's a really good point. I guess I'm abusing my role here as both a kernel developer who wants to see stable kernels released by the community, and as the Tumbleweed maintainer, sorry about that.
I'm all for helping kernel.org developers out, but at the expense of people who are counting on Tumbleweed to not mess their systems up, I probably shouldn't do that, as you and Lars and others point out.
What I'd suggest is some awareness campaign. With zypper, having Kernel:HEAD as a repository too is trivial, as is installing both kernels. "Want to help out? Boot into the kernel:head kernel and let us know how that goes." Systems that don't want to update but rather stick to the latest stable kernel - that is, my laptop ;-) - won't, but test environments might.
Maybe just have a Tumbleweed:Kernel repo that people can add to get this newer kernel if they want to be on even more of a bleading edge, that would like to the proper Kernel:HEAD release at the -rc5 time period. Perhaps that might balance the needs of users better?
Right, though I am not sure this is all that different from adding Kernel:HEAD as it is? Regards, Lars -- Architect Storage/HA SUSE LINUX Products GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 21284 (AG Nürnberg) "Experience is the name everyone gives to their mistakes." -- Oscar Wilde -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Wed, Aug 29, 2012 at 05:04:23PM +0200, Lars Marowsky-Bree wrote:
On 2012-08-28T06:46:46, Greg KH <gregkh@linux.com> wrote:
On Tue, Aug 28, 2012 at 08:55:56AM +0200, Mike Galbraith wrote:
In practice, rc5 vs .0 probably won't make much difference at all, but more users will be bitten, else the move was a failure. So this move explicitly plans on users being bitten, detracting from Tumbleweed's rolling ~stable allure, _evading_ developmental regression detection and reporting duties being a major attractor to running rolling ~stable ;-)
That's a really good point. I guess I'm abusing my role here as both a kernel developer who wants to see stable kernels released by the community, and as the Tumbleweed maintainer, sorry about that.
I'm all for helping kernel.org developers out, but at the expense of people who are counting on Tumbleweed to not mess their systems up, I probably shouldn't do that, as you and Lars and others point out.
What I'd suggest is some awareness campaign. With zypper, having Kernel:HEAD as a repository too is trivial, as is installing both kernels.
"Want to help out? Boot into the kernel:head kernel and let us know how that goes."
Systems that don't want to update but rather stick to the latest stable kernel - that is, my laptop ;-) - won't, but test environments might.
Maybe just have a Tumbleweed:Kernel repo that people can add to get this newer kernel if they want to be on even more of a bleading edge, that would like to the proper Kernel:HEAD release at the -rc5 time period. Perhaps that might balance the needs of users better?
Right, though I am not sure this is all that different from adding Kernel:HEAD as it is?
Kernel:HEAD gets -rc1 updates from what I remember (note, this could be wrong), so yes, it is almost like that. And, as you point out, it's so close, that it probably doesn't help out much by me doing the extra work to wait until -rc5 to add the package to a different repo. So, I think I'm going to back down from this proposal, and just leave things as they are. Jos and you and others rightly point out that Tumbleweed is supposed to be "stable", and by doing kernel updates at -rc5, just to get kernel testers to find problems, kind of goes against that original goal. So, thanks to all for the feedback, and for the sanity check, I'll just leave things as-is and not change the way the kernel is being updated. thanks, greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 8/29/12 1:58 PM, Greg KH wrote:
On Wed, Aug 29, 2012 at 05:04:23PM +0200, Lars Marowsky-Bree wrote:
On 2012-08-28T06:46:46, Greg KH <gregkh@linux.com> wrote:
On Tue, Aug 28, 2012 at 08:55:56AM +0200, Mike Galbraith wrote:
In practice, rc5 vs .0 probably won't make much difference at all, but more users will be bitten, else the move was a failure. So this move explicitly plans on users being bitten, detracting from Tumbleweed's rolling ~stable allure, _evading_ developmental regression detection and reporting duties being a major attractor to running rolling ~stable ;-)
That's a really good point. I guess I'm abusing my role here as both a kernel developer who wants to see stable kernels released by the community, and as the Tumbleweed maintainer, sorry about that.
I'm all for helping kernel.org developers out, but at the expense of people who are counting on Tumbleweed to not mess their systems up, I probably shouldn't do that, as you and Lars and others point out.
What I'd suggest is some awareness campaign. With zypper, having Kernel:HEAD as a repository too is trivial, as is installing both kernels.
"Want to help out? Boot into the kernel:head kernel and let us know how that goes."
Systems that don't want to update but rather stick to the latest stable kernel - that is, my laptop ;-) - won't, but test environments might.
Maybe just have a Tumbleweed:Kernel repo that people can add to get this newer kernel if they want to be on even more of a bleading edge, that would like to the proper Kernel:HEAD release at the -rc5 time period. Perhaps that might balance the needs of users better?
Right, though I am not sure this is all that different from adding Kernel:HEAD as it is?
Kernel:HEAD gets -rc1 updates from what I remember (note, this could be wrong), so yes, it is almost like that.
- -rc2 but I mentioned in another reply I'd have no problem pushing it at -rc1 instead if there's interest. I usually use -rc2 because there is a lot of stabilization that happens between the two. - -Jeff
And, as you point out, it's so close, that it probably doesn't help out much by me doing the extra work to wait until -rc5 to add the package to a different repo.
So, I think I'm going to back down from this proposal, and just leave things as they are. Jos and you and others rightly point out that Tumbleweed is supposed to be "stable", and by doing kernel updates at -rc5, just to get kernel testers to find problems, kind of goes against that original goal.
So, thanks to all for the feedback, and for the sanity check, I'll just leave things as-is and not change the way the kernel is being updated.
thanks,
greg k-h
- -- Jeff Mahoney SUSE Labs -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.18 (Darwin) Comment: GPGTools - http://gpgtools.org Comment: Using GnuPG with Mozilla - http://www.enigmail.net/ iQIcBAEBAgAGBQJQPozVAAoJEB57S2MheeWyzQwQAJRGaiqie7NaDzu6R9tg41yQ UUE6jBzrY/NTCVQ45GFrrn0004sZ7TW8ZgPfog7T5OI+3V6DTn6K3pgULztooj4t sGMgVr1pPWuRKl20WXt1bpCYca87GTzbbmVhFInoZYvHHnQGuiPW7r+ki+t5zhpV bxONBsAzOZckw+YahemHVbmboRT7uJeDtcNYxMrR/oSgdmH4wnH653NBikarbD5K pbdTO5XfT1oxGhyTYVdJZOx9HI8/+0GHCm7UpWas682Uz5XZQrz51x4cc/tUY6QL QN3dIM+gV4e9Xts5faeak6gzYJs6eayVR070n3ygMjisQS/YASTqnVTKAPBtD5Zb hX25FTY/L7HjxOx0IX7tg9QhiR1zu/Ov64+afBlMS0ZpROnyUSlTNC49ywKmrC2N bYrQr9kIGbMI13ljli2CHlnTQHdI162FjHfj88YY/wkBL9xzI7k2jLKjnjsG1aON OZPXJCDBRIz835w4CHz3Pw1785w4ymK8sNVsC/hp0AAA8seLxJOqtUvedn3a7zO7 pwaz7bDhgfMweEir2BO872x27Z13CQ6s4L5iYTDYXADhO077zx+InL6tkZ1cgeiS w6jU3uXNDD6fYxGAnnIOQCN8k+ncQPCsahpx1/+jSSJl4qqvadEamAPaytmFhBMg XDz80qOyNYrj/LkuK8lj =aeoi -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 08/28/2012 07:05 AM, Greg KH wrote:
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
Hi, I have no objections to that. However I'm not sure how do you want to do it technically. Are you going to create a new Kernel:* project for that purpose? Since I don't think it's a good idea to put -rc kernels into Kernel:stable (and I would actually object to this). thanks, - -- js suse labs -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.18 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://www.enigmail.net/ iQIcBAEBAgAGBQJQPG/JAAoJEL0lsQQGtHBJEEcP/AybfugavfYWq37QTjtaRDgR rFNxCwODR/QowVXSDR2KIr3jPGSaObsuFQm7DJIpcb06JtYzWvFjbHf5XDQC+zrO ZbV1J4ULMLCu8TN13wBvBH9N2HHIrk2cJLV8RpNaWC5zIcvAVeDk8Hn1U5EeP+fo jHQ42ZtDqmoPMSlURQo3hC3eFI+nmIQcN/1p6HhXT4qnmN+gpGrLO/qHOgZOlQJQ 84uTtS9XsDyzEY/l0dxFawAUoyvjVCK2KRXw2IgGHQ8Jt8bsLpp0z8TOHq6qNyDz EN4NGfuV6Gj719/PgH7HUWr8xNpkieDerAX1WoDJqkpMnQGA5EzOdXi6PsGSd7ka abZqe4E/S2ZrKr8pFrQodCShGQgzCWek4ouMXe4VHevWe7Wi6zi8JZKzQ3BWoyv7 GSr2+rwt+XYbGcf1pot408H0eRmJzMhQRA7BAVOWNPp7oJTEJ5F1Z2NHFJD0k9vP hvO7j0gui+sjFTEmc5YGpNlCdAQrHj5xGtTaLl1ccec1r+DjhT4ToO17/E6mQlLS 6Eru4/fGexQJ4f5zM/sgYbgSBKe1lNDg/VrOpc6Fgl4hFrUNf/yc3PKHnX+C/F6n XlbtRN+8bhrcgkOaQpWxpknh2z1qJIMgpgTVgbDYEKD0atgziVgOHcvSBddjPPs5 VpJCDBZFyidp1k3CaDse =Ul03 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Tue, Aug 28, 2012 at 09:14:17AM +0200, Jiri Slaby wrote:
On 08/28/2012 07:05 AM, Greg KH wrote:
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
Hi, I have no objections to that. However I'm not sure how do you want to do it technically. Are you going to create a new Kernel:* project for that purpose? Since I don't think it's a good idea to put -rc kernels into Kernel:stable (and I would actually object to this).
No, I would be linking to the kernels in Kernel:HEAD that Jeff creates, not create a new one. thanks, greg k-h -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On 2012-08-27T22:05:23, Greg KH <gregkh@linux.com> wrote: Hi Greg, I'm not a huge fan of this. I use Tumbleweed as "rolling stable". I get my beta kernels from Kernel:HEAD. I don't think pushing beta/rcs to Tumbleweed is a good idea; it mixes "rolling updades" with "hey, how about you become a beta tester?". It's not a strong objection, since I have multiple kernels installed anyway; but it strikes me as inconsistent with the Tumbleweed goals (as I understand them). Regards, Lars -- Architect Storage/HA SUSE LINUX Products GmbH, GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer, HRB 21284 (AG Nürnberg) "Experience is the name everyone gives to their mistakes." -- Oscar Wilde -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
Hi, On Tue, Aug 28, 2012 at 11:10:51AM +0200, Lars Marowsky-Bree wrote:
I use Tumbleweed as "rolling stable". I get my beta kernels from Kernel:HEAD.
I don't think pushing beta/rcs to Tumbleweed is a good idea; it mixes "rolling updades" with "hey, how about you become a beta tester?".
I think the same. I use Tumbleweed on some machines because I know the risk is higher than with openSUSE x.x, but first it is lower than with Factory and second it seems I can handle most of the issues myself. Additionally I use it because Greg is a kernel guy and knows a lot about the kernel he checks into Tumbleweed. Maybe this part was a wrong assumption :) However, having a more bleeding edge kernel would scare me away. At least that was my first reaction. I don't update very often, so I would running into a higher risk to have a kernel which is buggy. Enforcing upgrades is meaningful for Factory, but not really from what I thought Tumbleweed was about. Maybe it would be better to have a general kernel-latest, kernel-beta or kernel-rc instead of changing the default kernel for a certain codestream. -- Bye, Stephan Barth SUSE Technical Services - SUSE LINUX GmbH GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer HRB 21284 (AG Nürnberg) Maxfeldstr. 5, D-90409 Nuremberg Register at suse.com/susecon - Follow us at twitter.com/susecon12 -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 8/28/12 1:05 AM, Greg KH wrote:
Hi all,
As part of one of the Linux kernel Summit discussions today, it was brought up that after a kernel is released (for example 3.5), it's a bit too late to be doing testing to see how well it is working out. The .0 release is usually a bit rough, and it takes until the .1 or .2 release to get most major issues out of the way.
So, the kernel developers would like to get a wider range of testing, and one thing proposed would be to have rolling distros switch their kernel over a bit earlier to the new release than they had been doing. Specifically, around the -rc5 point in time would be great. That way, any reported regressions could be fixed sooner and get into the final .0 release for everyone to use.
Now this does place a bit of a larger burden on the users of those distros to be diligent in reporting problems, and the distro engineers to report the issues upstream as well, but it sounds like a reasonable thing to try out.
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
- From a technical perspective, it doesn't make much difference to me. The first step in triaging non-obvious Tumbleweed kernel bugs is to ask that the reporter reproduce with a kernel-vanilla from the same vintage anyway. Unless effort comes in from elsewhere, Tumbleweed kernels will never get as much attention as release kernels. However, and several other replies have noted this as well, it seems that this strategy is conflating the goals of the upstream kernel community with those of the Tumbleweed project. The benefit to the upstream kernel community is obvious -- but I'm not seeing the up-side for Tumbleweed users. Tumbleweed *isn't* Factory, nor should it be. I think you'll find that if the kernel becomes even less stable in Tumbleweed, users will either stop using it or start marking the kernel packages as Taboo for update until a .0 is released. If the goal is test coverage, I think the best way to do that is to make pre-built kernels available for users who want to participate in that process. Co-opting Tumbleweed isn't the way to do that. We already have Kernel:HEAD for that and I sync that starting with - -rc2. I've (we've?) have that policy for years, but now that Jiri has been maintaining Kernel:stable for a while, it may be time to make Kernel:HEAD even more aggressive and start with -rc1. Sure, it can lag a little behind for things like Xen, but the upstream testing community doesn't care about that anyway. Perhaps the answer is to come up with a way to allow users to ride the bleeding edge without having to hunt down OBS published repo URLs themselves. FWIW, using -rc1 in HEAD doesn't increase my work effort much. It would probably lessen it, even, since I merge -rc1 as soon as it's released and then update to -rc2 before pulling into master anyway. - -Jeff - -- Jeff Mahoney SUSE Labs -----BEGIN PGP SIGNATURE----- Version: GnuPG/MacGPG2 v2.0.18 (Darwin) Comment: GPGTools - http://gpgtools.org Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQIcBAEBAgAGBQJQPMSRAAoJEB57S2MheeWyN8IP/jX7z//X3UYxGGaQBy0d5djy i3HaTTFX84SD34oGj/qINSdF853HwRWJbkwGIGue0cRmAtcM1nEgWIB9aiXbKlHR vTMmSXLhGRtPNz5wCBiCswSzQwy6XD7D6zbqrarOk356GngDAK5lGOVIsA3lllo/ lFE4otWqQ3KNUlFWxi7wsiuztyvlqyvLpojxDWknf40Xzz+CojNuR0tPay/87s5n C8QQzxoKPodIwAgn1DC6AmZDpFToFN/qhAEb4V+Fo9YlQosO+ljfiD7SdaO3Jr1n m9/bV2kBGeW/dgrKVL4iZf+FpjvqO8CCVA/Y8PwogWCTJu4tr1t4Bjal7jJHjdN+ 5EYEFkXmYXSU9X2YD2j5+sqPgZ3y6eu22wfqi8cVS2B7vBX70jGKyeUQ+Uw0e3hu rSVYL71EpB/L6s9GOrhqKmDfA+Be8sCoZBRI8Igzb8raWCdoI9DyHhL9EMTHheH0 ShvMbrfpLLMpfbG30SUSAgL5+RIHViDmmg8ORBsrsq5Fkmh7LkmwxFUdNqx5lGW8 T/UpJHWd5C79BCOBaF5VMSqb3K1X6RalW6VYQZnm1NWPv2m0vEy62VgP5q89LDTV ffwmSQj/CIZRg18D9P07tjVuligsyErc1FgTLrLQ6wzgC9A3X63qKl2KjFKVNkNv ol8PEMnNIaGczze7XjHg =H4IL -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse-kernel+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kernel+owner@opensuse.org
On Monday, August 27, 2012 22:05:23 Greg KH wrote:
Hi all,
As part of one of the Linux kernel Summit discussions today, it was brought up that after a kernel is released (for example 3.5), it's a bit too late to be doing testing to see how well it is working out. The .0 release is usually a bit rough, and it takes until the .1 or .2 release to get most major issues out of the way.
So, the kernel developers would like to get a wider range of testing, and one thing proposed would be to have rolling distros switch their kernel over a bit earlier to the new release than they had been doing. Specifically, around the -rc5 point in time would be great. That way, any reported regressions could be fixed sooner and get into the final .0 release for everyone to use.
Now this does place a bit of a larger burden on the users of those distros to be diligent in reporting problems, and the distro engineers to report the issues upstream as well, but it sounds like a reasonable thing to try out.
So, for the next kernel release, I'm thinking of switching the kernel in Tumbleweed over to 3.6 at the -rc5 timeframe. Does anyone strongly object to this happening?
Actually, I'd personally prefer to move to a new kernel in the .1 times instead - have Tumbleweed MORE stable instead of less. Rationale: IF we take Tumbleweed as a prominent way for people to get newer oS software between releases (especially important if we lengthen the release cycle, which is at least on the agenda), Tumbleweed should become MORE, not less stable. Factory would be where RC's of kernels get tested, and so would a possible tumbleweed-testing repository. For me as end user, I don't use Tumbleweed for testing (that's what Factory is for, imho) but to have newer software without waiting 8 months. Making it less stable would make me leave Tumbleweed and run the latest openSUSE with a large number of OBS repo's again (as I did before Tumbleweed). But this choice comes down to a philosophical question: what's tumbleweed for? Is it (as we currently advertise) a STABLE rolling release of openSUSE or is it a testing ground for new software? My choice would be obvious from what I wrote. Note that imho the choice of what Tumbleweed is for lies with the Tumbleweed maintainer - that's you. But I'd highly recommend changing our promotion of Tumbleweed if you do this (and similar steps in the direction of a more Factory-like Tumbleweed): it should no longer be promoted to our average user. And, if this is done, I'd advocate for either shortening or at least NOT lengthening our release schedule, OR creating a new, STABLE tumbleweed branch which IS meant for end users. /Jos
thanks,
greg k-h
participants (9)
-
Andrew Wafaa -
Greg KH -
Jeff Mahoney -
Jiri Slaby -
Jos Poortvliet -
Lars Marowsky-Bree -
M. Edward (Ed) Borasky -
Mike Galbraith -
Stephan Barth