| Message ID | cover.1709541697.git.zhengqi.arch@bytedance.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | minor fixes and supplement for ptdesc | expand |
On Mon, Mar 04, 2024 at 07:07:17PM +0800, Qi Zheng wrote: > Hi all, Sorry for the late review. Thanks for looking at doing some ptdesc conversions. This patchset has the right idea and looks *mostly* fine. > In this series, the [PATCH 1/3] and [PATCH 2/3] are fixes for some issues > discovered during code inspection. > > The [PATCH 3/3] is a supplement to ptdesc conversion in s390, I don't know > why this is not done in the commit 6326c26c1514 ("s390: convert various pgalloc > functions to use ptdescs"), maybe I missed something. And since I don't have an It's important to keep in mind the end goal of ptdescs, cleaning up much of the struct page field misuse by standardizing their usage. s390 page tables and gmap are similar but not the same, so the conversions require deeper thought. My initial "Split ptdesc from struct page" patchset tried to focus on the most straightforward, simple conversions in order to introduce the descriptor and lay a foundation for future conversions - you can see some more complicated iterations v6 and prior. When converting to ptdescs (and any other newer descriptors), we should be careful about generating superficial code churn instead of using them to solve the problems they are trying to solve.
On 26.03.24 20:07, Vishal Moola wrote: > On Mon, Mar 04, 2024 at 07:07:17PM +0800, Qi Zheng wrote: >> Hi all, > > Sorry for the late review. Thanks for looking at doing some ptdesc > conversions. This patchset has the right idea and looks *mostly* fine. > >> In this series, the [PATCH 1/3] and [PATCH 2/3] are fixes for some issues >> discovered during code inspection. >> >> The [PATCH 3/3] is a supplement to ptdesc conversion in s390, I don't know >> why this is not done in the commit 6326c26c1514 ("s390: convert various pgalloc >> functions to use ptdescs"), maybe I missed something. And since I don't have an > > It's important to keep in mind the end goal of ptdescs, cleaning up much > of the struct page field misuse by standardizing their usage. s390 page > tables and gmap are similar but not the same, so the conversions require > deeper thought. > > My initial "Split ptdesc from struct page" patchset tried to focus on the > most straightforward, simple conversions in order to introduce the > descriptor and lay a foundation for future conversions - you can see some > more complicated iterations v6 and prior. > > When converting to ptdescs (and any other newer descriptors), we should > be careful about generating superficial code churn instead of using > them to solve the problems they are trying to solve. The gmap shadow pages are page tables that are not linked into the user page tables. I recall I raised in the past that using ptdesc from them is confusing.
Hi all, In this series, the [PATCH 1/3] and [PATCH 2/3] are fixes for some issues discovered during code inspection. The [PATCH 3/3] is a supplement to ptdesc conversion in s390, I don't know why this is not done in the commit 6326c26c1514 ("s390: convert various pgalloc functions to use ptdescs"), maybe I missed something. And since I don't have an s390 environment, I hope kernel test robot can help compile and test, and this is why I did not fold [PATCH 2/3] and [PATCH 3/3] into one patch. Comments and suggestions are welcome. Thanks, Qi Qi Zheng (3): mm: pgtable: correct the wrong comment about ptdesc->__page_flags mm: pgtable: add missing pt_index to struct ptdesc s390: supplement for ptdesc conversion arch/s390/include/asm/pgalloc.h | 4 ++-- arch/s390/mm/gmap.c | 38 +++++++++++++++++---------------- arch/s390/mm/pgalloc.c | 8 +++---- include/linux/mm_types.h | 5 ++++- 4 files changed, 30 insertions(+), 25 deletions(-)