Hi Amit,
since I've got a backport of EASv3 and I'd like to share this with all, can you please give me an update about the EAS backport tree maintenance bits we agreed upon.
Thanks a lot,
- Juri
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Regards, Amit [1] Since we don't _yet_ have the complex interplay between features, we decided at Connect to lower the barrier for new features until we have a full validation cycle working and putting out reports. By then we'll need to agree on the acceptance criteria for letting in new features.
On Wed, Feb 18, 2015 at 8:05 PM, Juri Lelli juri.lelli@arm.com wrote:
Hi Amit,
since I've got a backport of EASv3 and I'd like to share this with all, can you please give me an update about the EAS backport tree maintenance bits we agreed upon.
Thanks a lot,
- Juri
Hi Amit,
On 19/02/15 10:40, Amit Kucheria wrote:
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Thanks for the information, I've found the new branches and used them as a basis for the new backport. However, I also noticed that the wikipage is not yet updated. Can we put some information on how we intend to manage the thing there?
Regarding the EASv3 backport, I had also to backport 5 additional patches from mainline that should go on top of the upstream branch. The EASv3 backport patches apply on top of upstream (with those 5 patches). I attach an archive containing them. Is there a proper way to issue a pull request to Alex or is this way just fine? I could post them on the list as a patch series maybe.
Thanks a lot.
Best,
- Juri
Regards, Amit [1] Since we don't _yet_ have the complex interplay between features, we decided at Connect to lower the barrier for new features until we have a full validation cycle working and putting out reports. By then we'll need to agree on the acceptance criteria for letting in new features.
On Wed, Feb 18, 2015 at 8:05 PM, Juri Lelli juri.lelli@arm.com wrote:
Hi Amit,
since I've got a backport of EASv3 and I'd like to share this with all, can you please give me an update about the EAS backport tree maintenance bits we agreed upon.
Thanks a lot,
- Juri
On 02/27/2015 08:17 PM, Juri Lelli wrote:
Hi Amit,
On 19/02/15 10:40, Amit Kucheria wrote:
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Thanks for the information, I've found the new branches and used them as a basis for the new backport. However, I also noticed that the wikipage is not yet updated. Can we put some information on how we intend to manage the thing there?
Regarding the EASv3 backport, I had also to backport 5 additional patches from mainline that should go on top of the upstream branch. The EASv3 backport patches apply on top of upstream (with those 5 patches). I attach an archive containing them. Is there a proper way to issue a pull request to Alex or is this way just fine? I could post them on the list as a patch series maybe.
A pull request would make merge more easy for me.
-- Thanks Alex
Hi Alex,
On 28/02/15 07:50, Alex Shi wrote:
On 02/27/2015 08:17 PM, Juri Lelli wrote:
Hi Amit,
On 19/02/15 10:40, Amit Kucheria wrote:
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Thanks for the information, I've found the new branches and used them as a basis for the new backport. However, I also noticed that the wikipage is not yet updated. Can we put some information on how we intend to manage the thing there?
Regarding the EASv3 backport, I had also to backport 5 additional patches from mainline that should go on top of the upstream branch. The EASv3 backport patches apply on top of upstream (with those 5 patches). I attach an archive containing them. Is there a proper way to issue a pull request to Alex or is this way just fine? I could post them on the list as a patch series maybe.
A pull request would make merge more easy for me.
I managed to setup two externally accessible branches from which you should be able to pull the EASv3 code update.
repo: git://linux-arm.org/linux-power.git
branches: linaro_upstream/linaro-sched-upstream-forAlex [c22b383..HEAD] linaro_upstream/linaro-sched-upstream-easv3-forAlex [83ad17f..HEAD]
Please let me know if everything works fine.
Best Regards,
- Juri
On 03/16/2015 08:18 PM, Juri Lelli wrote:
I managed to setup two externally accessible branches from which you should be able to pull the EASv3 code update.
repo: git://linux-arm.org/linux-power.git
branches: linaro_upstream/linaro-sched-upstream-forAlex [c22b383..HEAD] linaro_upstream/linaro-sched-upstream-easv3-forAlex [83ad17f..HEAD]
Please let me know if everything works fine.
pulled, both of them works good. Thanks!
- Thanks Alex
On 19/03/15 02:43, Alex Shi wrote:
On 03/16/2015 08:18 PM, Juri Lelli wrote:
I managed to setup two externally accessible branches from which you should be able to pull the EASv3 code update.
repo: git://linux-arm.org/linux-power.git
branches: linaro_upstream/linaro-sched-upstream-forAlex [c22b383..HEAD] linaro_upstream/linaro-sched-upstream-easv3-forAlex [83ad17f..HEAD]
Please let me know if everything works fine.
pulled, both of them works good. Thanks!
Cool! Thanks a lot.
Best,
- Juri
Thanks Alex
On Fri, Feb 27, 2015 at 5:47 PM, Juri Lelli juri.lelli@arm.com wrote:
Hi Amit,
On 19/02/15 10:40, Amit Kucheria wrote:
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Thanks for the information, I've found the new branches and used them as a basis for the new backport. However, I also noticed that the wikipage is not yet updated. Can we put some information on how we intend to manage the thing there?
I was hoping to do a couple of rounds of merging new code in to iron our any kinks in the process before publishing the tree maintenance process. Besides, IMHO the document we currently have is too detailed and targeted to contributors to the code line, not to the users who just want to take the branches and test it. Do you agree?
The other thing that is still left unresolved is what will be the acceptance criteria for patches in the future? Here are a few different vectors: - Benchmark improvements for a defined set of benchmarks (Start with WAv2 bundled-benchmarks?) - Ack from other contributors - What test platform to consider as representative for our testing (Juno?)
Regarding the EASv3 backport, I had also to backport 5 additional patches from mainline that should go on top of the upstream branch. The EASv3 backport patches apply on top of upstream (with those 5 patches). I attach an archive containing them. Is there a proper way to issue a pull request to Alex or is this way just fine? I could post them on the list as a patch series maybe.
As Alex suggested, let us use well-known git workflows. In this case, I'd send two pull-requests to eas-dev and Alex - one to update the sched-upstream branch, the other to upstream sched-core. There should also be some comments on what sort of testing has been done with these patches. I believe that eventually you and Alex will be the people to decide what code should be merged and what should be rejected.
Thanks a lot.
Best,
- Juri
Regards, Amit
On 03/02/2015 02:34 PM, Amit Kucheria wrote:
On Fri, Feb 27, 2015 at 5:47 PM, Juri Lelli juri.lelli@arm.com wrote:
Hi Amit,
On 19/02/15 10:40, Amit Kucheria wrote:
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Thanks for the information, I've found the new branches and used them as a basis for the new backport. However, I also noticed that the wikipage is not yet updated. Can we put some information on how we intend to manage the thing there?
I was hoping to do a couple of rounds of merging new code in to iron our any kinks in the process before publishing the tree maintenance process. Besides, IMHO the document we currently have is too detailed and targeted to contributors to the code line, not to the users who just want to take the branches and test it. Do you agree?
Why not fill both info on the wiki page, put them on different part?
The other thing that is still left unresolved is what will be the acceptance criteria for patches in the future? Here are a few different vectors:
- Benchmark improvements for a defined set of benchmarks (Start with
WAv2 bundled-benchmarks?)
That can show the benefit of this eas branch and persude more member to try it.
- Ack from other contributors
That would be great to involve more review!
- What test platform to consider as representative for our testing (Juno?)
Any other big.little platform which you'd like to suggest, Amit?
Regarding the EASv3 backport, I had also to backport 5 additional patches from mainline that should go on top of the upstream branch. The EASv3 backport patches apply on top of upstream (with those 5 patches). I attach an archive containing them. Is there a proper way to issue a pull request to Alex or is this way just fine? I could post them on the list as a patch series maybe.
As Alex suggested, let us use well-known git workflows. In this case, I'd send two pull-requests to eas-dev and Alex - one to update the sched-upstream branch, the other to upstream sched-core. There should also be some comments on what sort of testing has been done with these patches. I believe that eventually you and Alex will be the people to decide what code should be merged and what should be rejected.
Thanks a lot.
Best,
- Juri
Regards, Amit
-- Thanks Alex
On 02/27/2015 08:17 PM, Juri Lelli wrote:
Hi Amit,
On 19/02/15 10:40, Amit Kucheria wrote:
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Thanks for the information, I've found the new branches and used them as a basis for the new backport. However, I also noticed that the wikipage is not yet updated. Can we put some information on how we intend to manage the thing there?
Regarding the EASv3 backport, I had also to backport 5 additional patches from mainline that should go on top of the upstream branch. The EASv3 backport patches apply on top of upstream (with those 5 patches). I attach an archive containing them. Is there a proper way to issue a pull request to Alex or is this way just fine? I could post them on the list as a patch series maybe.
Hi, Juri the upstream branch is ok for me. but when applying the easv6 by 'git am' basing on new upstream branch, I hit the following error: Would you like to look into this issue?
Applying: sched: Infrastructure to query if load balancing is energy-aware error: patch failed: kernel/sched/fair.c:5536 error: kernel/sched/fair.c: patch does not apply Patch failed at 0037 sched: Infrastructure to query if load balancing is energy-aware The copy of the patch that failed is found in: /home/alexs/lsk/kernel/.git/rebase-apply/patch When you have resolved this problem, run "git am --continue". If you prefer to skip this patch, run "git am --skip" instead. To restore the original branch and stop patching, run "git am --abort".
Hi Alex,
On 02/03/15 12:23, Alex Shi wrote:
On 02/27/2015 08:17 PM, Juri Lelli wrote:
Hi Amit,
On 19/02/15 10:40, Amit Kucheria wrote:
Hi Juri,
Not sure what update you were waiting for, but Alex has gone ahead and split the current patch stack into stable/sched-upstream and stable/sched-core branches. So any updates from you should apply on top of the stable/sched-core branch. If you've rebased the patchset, please let Alex know in your pull request so we can replace the entire branch[1].
Thanks for the information, I've found the new branches and used them as a basis for the new backport. However, I also noticed that the wikipage is not yet updated. Can we put some information on how we intend to manage the thing there?
Regarding the EASv3 backport, I had also to backport 5 additional patches from mainline that should go on top of the upstream branch. The EASv3 backport patches apply on top of upstream (with those 5 patches). I attach an archive containing them. Is there a proper way to issue a pull request to Alex or is this way just fine? I could post them on the list as a patch series maybe.
Hi, Juri the upstream branch is ok for me. but when applying the easv6 by 'git am' basing on new upstream branch, I hit the following error: Would you like to look into this issue?
The two patchsets seem to apply fine one on top of the other to me. Could you please check that you have Kirill's "sched/fair: Remove double_lock_balance() from load_balance()" in your tree?
Anyway, I'll try to setup an externally accessible repo for a proper pull request.
Best,
- Juri
Applying: sched: Infrastructure to query if load balancing is energy-aware error: patch failed: kernel/sched/fair.c:5536 error: kernel/sched/fair.c: patch does not apply Patch failed at 0037 sched: Infrastructure to query if load balancing is energy-aware The copy of the patch that failed is found in: /home/alexs/lsk/kernel/.git/rebase-apply/patch When you have resolved this problem, run "git am --continue". If you prefer to skip this patch, run "git am --skip" instead. To restore the original branch and stop patching, run "git am --abort".
-
Alex Shi -
Amit Kucheria -
Juri Lelli