Re: Process identical patches in different tree
On 09/01/2020 02:53, CK Hu wrote: > Hi, Matthias: > > On Wed, 2020-01-08 at 13:05 +0100, Matthias Brugger wrote: >> On 08/01/2020 12:14, Matthias Brugger wrote: >>> Hi CK, >>> >>> On 07/01/2020 03:56, CK Hu wrote: Hi, Dave, Daniel, Matthias: In mediatek-drm-next-5.6 [1], I've cherry-pick 3 patches from v5.5-next/soc [2] because some drm patches depend on these cmdq patches. So these cmdq patches exist in both tree now. I want to know how to process this case. I think we could choose one of below way: 1. Because these cmdq patches are identical in both tree, so each tree could do its own upstream and the there would be nothing happen when merge. 2. Let soc upstream first, and mediatek drm rebase on the latest mainline then upstream. Which one do you prefer? >>> >>> What we would need is a stable branch with this commits that get merged by >>> both >>> trees. If I understand correctly that otherwise the SHA of the commits >>> would be >>> different and that would provoke merge conflicts. >>> >>> We should not rely on one tree being merged before the other. AFAIK there >>> is no >>> hard merge order between trees. >>> >> >> I prepared a branch with the patches I think are relevant for you. Please >> confirm that this is correct, merge the tree in yours and I'll do the same >> for >> v5.5-next/soc >> >> >> >> The following changes since commit e42617b825f8073569da76dc4510bfa019b1c35a: >> >> Linux 5.5-rc1 (2019-12-08 14:57:55 -0800) >> >> are available in the Git repository at: >> >> https://git.kernel.org/pub/scm/linux/kernel/git/matthias.bgg/linux.git/ >> tags/v5.5-next-cmdq-stable >> >> for you to fetch changes up to d412f18c9bc791d8951e903de9a68817e3098a6a: >> >> soc: mediatek: cmdq: add cmdq_dev_get_client_reg function (2020-01-08 >> 12:59:57 >> +0100) >> >> >> cmdq patches needed by drm driver to use cmdq interface >> >> >> Bibby Hsieh (4): >> soc: mediatek: cmdq: remove OR opertaion from err return >> soc: mediatek: cmdq: define the instruction struct >> soc: mediatek: cmdq: add polling function >> soc: mediatek: cmdq: add cmdq_dev_get_client_reg function >> >> drivers/soc/mediatek/mtk-cmdq-helper.c | 147 >> - >> include/linux/mailbox/mtk-cmdq-mailbox.h | 11 ++ >> include/linux/soc/mediatek/mtk-cmdq.h| 53 + >> 3 files changed, 185 insertions(+), 26 deletions(-) >> >> >> > > I've done in [1], is it what you expect? > > [1] > https://github.com/ckhu-mediatek/linux.git-tags/commits/mediatek-drm-next-5.6 > Looks good to me :) Regards, Matthias ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: Process identical patches in different tree
Hi, Matthias: On Wed, 2020-01-08 at 13:05 +0100, Matthias Brugger wrote: > On 08/01/2020 12:14, Matthias Brugger wrote: > > Hi CK, > > > > On 07/01/2020 03:56, CK Hu wrote: > >> Hi, Dave, Daniel, Matthias: > >> > >> In mediatek-drm-next-5.6 [1], I've cherry-pick 3 patches from > >> v5.5-next/soc [2] because some drm patches depend on these cmdq patches. > >> So these cmdq patches exist in both tree now. I want to know how to > >> process this case. I think we could choose one of below way: > >> > >> 1. Because these cmdq patches are identical in both tree, so each tree > >> could do its own upstream and the there would be nothing happen when > >> merge. > >> 2. Let soc upstream first, and mediatek drm rebase on the latest > >> mainline then upstream. > >> > >> Which one do you prefer? > >> > > > > What we would need is a stable branch with this commits that get merged by > > both > > trees. If I understand correctly that otherwise the SHA of the commits > > would be > > different and that would provoke merge conflicts. > > > > We should not rely on one tree being merged before the other. AFAIK there > > is no > > hard merge order between trees. > > > > I prepared a branch with the patches I think are relevant for you. Please > confirm that this is correct, merge the tree in yours and I'll do the same for > v5.5-next/soc > > > > The following changes since commit e42617b825f8073569da76dc4510bfa019b1c35a: > > Linux 5.5-rc1 (2019-12-08 14:57:55 -0800) > > are available in the Git repository at: > > https://git.kernel.org/pub/scm/linux/kernel/git/matthias.bgg/linux.git/ > tags/v5.5-next-cmdq-stable > > for you to fetch changes up to d412f18c9bc791d8951e903de9a68817e3098a6a: > > soc: mediatek: cmdq: add cmdq_dev_get_client_reg function (2020-01-08 > 12:59:57 > +0100) > > > cmdq patches needed by drm driver to use cmdq interface > > > Bibby Hsieh (4): > soc: mediatek: cmdq: remove OR opertaion from err return > soc: mediatek: cmdq: define the instruction struct > soc: mediatek: cmdq: add polling function > soc: mediatek: cmdq: add cmdq_dev_get_client_reg function > > drivers/soc/mediatek/mtk-cmdq-helper.c | 147 > - > include/linux/mailbox/mtk-cmdq-mailbox.h | 11 ++ > include/linux/soc/mediatek/mtk-cmdq.h| 53 + > 3 files changed, 185 insertions(+), 26 deletions(-) > > > I've done in [1], is it what you expect? [1] https://github.com/ckhu-mediatek/linux.git-tags/commits/mediatek-drm-next-5.6 Regards, CK > Regards, > Matthias > > ___ > Linux-mediatek mailing list > linux-media...@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-mediatek ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: Process identical patches in different tree
On 08/01/2020 12:14, Matthias Brugger wrote: > Hi CK, > > On 07/01/2020 03:56, CK Hu wrote: >> Hi, Dave, Daniel, Matthias: >> >> In mediatek-drm-next-5.6 [1], I've cherry-pick 3 patches from >> v5.5-next/soc [2] because some drm patches depend on these cmdq patches. >> So these cmdq patches exist in both tree now. I want to know how to >> process this case. I think we could choose one of below way: >> >> 1. Because these cmdq patches are identical in both tree, so each tree >> could do its own upstream and the there would be nothing happen when >> merge. >> 2. Let soc upstream first, and mediatek drm rebase on the latest >> mainline then upstream. >> >> Which one do you prefer? >> > > What we would need is a stable branch with this commits that get merged by > both > trees. If I understand correctly that otherwise the SHA of the commits would > be > different and that would provoke merge conflicts. > > We should not rely on one tree being merged before the other. AFAIK there is > no > hard merge order between trees. > I prepared a branch with the patches I think are relevant for you. Please confirm that this is correct, merge the tree in yours and I'll do the same for v5.5-next/soc The following changes since commit e42617b825f8073569da76dc4510bfa019b1c35a: Linux 5.5-rc1 (2019-12-08 14:57:55 -0800) are available in the Git repository at: https://git.kernel.org/pub/scm/linux/kernel/git/matthias.bgg/linux.git/ tags/v5.5-next-cmdq-stable for you to fetch changes up to d412f18c9bc791d8951e903de9a68817e3098a6a: soc: mediatek: cmdq: add cmdq_dev_get_client_reg function (2020-01-08 12:59:57 +0100) cmdq patches needed by drm driver to use cmdq interface Bibby Hsieh (4): soc: mediatek: cmdq: remove OR opertaion from err return soc: mediatek: cmdq: define the instruction struct soc: mediatek: cmdq: add polling function soc: mediatek: cmdq: add cmdq_dev_get_client_reg function drivers/soc/mediatek/mtk-cmdq-helper.c | 147 - include/linux/mailbox/mtk-cmdq-mailbox.h | 11 ++ include/linux/soc/mediatek/mtk-cmdq.h| 53 + 3 files changed, 185 insertions(+), 26 deletions(-) Regards, Matthias ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: Process identical patches in different tree
Hi CK, On 07/01/2020 03:56, CK Hu wrote: > Hi, Dave, Daniel, Matthias: > > In mediatek-drm-next-5.6 [1], I've cherry-pick 3 patches from > v5.5-next/soc [2] because some drm patches depend on these cmdq patches. > So these cmdq patches exist in both tree now. I want to know how to > process this case. I think we could choose one of below way: > > 1. Because these cmdq patches are identical in both tree, so each tree > could do its own upstream and the there would be nothing happen when > merge. > 2. Let soc upstream first, and mediatek drm rebase on the latest > mainline then upstream. > > Which one do you prefer? > What we would need is a stable branch with this commits that get merged by both trees. If I understand correctly that otherwise the SHA of the commits would be different and that would provoke merge conflicts. We should not rely on one tree being merged before the other. AFAIK there is no hard merge order between trees. Regards, Matthias > [1] > https://github.com/ckhu-mediatek/linux.git-tags/commits/mediatek-drm-next-5.6 > [2] > https://git.kernel.org/pub/scm/linux/kernel/git/matthias.bgg/linux.git/log/?h=v5.5-next/soc > > Regards, > CK > ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Process identical patches in different tree
Hi, Dave, Daniel, Matthias: In mediatek-drm-next-5.6 [1], I've cherry-pick 3 patches from v5.5-next/soc [2] because some drm patches depend on these cmdq patches. So these cmdq patches exist in both tree now. I want to know how to process this case. I think we could choose one of below way: 1. Because these cmdq patches are identical in both tree, so each tree could do its own upstream and the there would be nothing happen when merge. 2. Let soc upstream first, and mediatek drm rebase on the latest mainline then upstream. Which one do you prefer? [1] https://github.com/ckhu-mediatek/linux.git-tags/commits/mediatek-drm-next-5.6 [2] https://git.kernel.org/pub/scm/linux/kernel/git/matthias.bgg/linux.git/log/?h=v5.5-next/soc Regards, CK ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel