From: "Ma, Jianpeng" <jianpeng.ma@intel.com>
To: "yukuai (C)" <yukuai3@huawei.com>, "Li, Coly" <colyli@suse.de>,
"axboe@kernel.dk" <axboe@kernel.dk>
Cc: "linux-bcache@vger.kernel.org" <linux-bcache@vger.kernel.org>,
"linux-block@vger.kernel.org" <linux-block@vger.kernel.org>,
"Ren, Qiaowei" <qiaowei.ren@intel.com>,
Christoph Hellwig <hch@lst.de>,
"Williams, Dan J" <dan.j.williams@intel.com>,
Hannes Reinecke <hare@suse.de>
Subject: RE: [PATCH v13 05/12] bcache: bch_nvmpg_free_pages() of the buddy allocator
Date: Tue, 22 Feb 2022 05:03:51 +0000 [thread overview]
Message-ID: <BN7PR11MB270659A970DFD7E01F2D4A38FD3B9@BN7PR11MB2706.namprd11.prod.outlook.com> (raw)
In-Reply-To: <e1e08398-dccb-6c0d-aaa9-da72f0cf9ef1@huawei.com>
> -----Original Message-----
> From: yukuai (C) <yukuai3@huawei.com>
> Sent: Monday, February 21, 2022 8:36 PM
> To: Li, Coly <colyli@suse.de>; axboe@kernel.dk
> Cc: linux-bcache@vger.kernel.org; linux-block@vger.kernel.org; Ma, Jianpeng
> <jianpeng.ma@intel.com>; Ren, Qiaowei <qiaowei.ren@intel.com>; Christoph
> Hellwig <hch@lst.de>; Williams, Dan J <dan.j.williams@intel.com>; Hannes
> Reinecke <hare@suse.de>
> Subject: Re: [PATCH v13 05/12] bcache: bch_nvmpg_free_pages() of the buddy
> allocator
>
> 在 2021/12/13 1:05, Coly Li 写道:
> > From: Jianpeng Ma <jianpeng.ma@intel.com>
> >
> > This patch implements the bch_nvmpg_free_pages() of the buddy allocator.
> >
> > The difference between this and page-buddy-free:
> > it need owner_uuid to free owner allocated pages, and must persistent
> > after free.
> >
> > Signed-off-by: Jianpeng Ma <jianpeng.ma@intel.com>
> > Co-developed-by: Qiaowei Ren <qiaowei.ren@intel.com>
> > Signed-off-by: Qiaowei Ren <qiaowei.ren@intel.com>
> > Cc: Christoph Hellwig <hch@lst.de>
> > Cc: Dan Williams <dan.j.williams@intel.com>
> > Cc: Hannes Reinecke <hare@suse.de>
> > Cc: Jens Axboe <axboe@kernel.dk>
> > ---
> > drivers/md/bcache/nvmpg.c | 164
> ++++++++++++++++++++++++++++++++++++--
> > drivers/md/bcache/nvmpg.h | 3 +
> > 2 files changed, 160 insertions(+), 7 deletions(-)
> >
> > diff --git a/drivers/md/bcache/nvmpg.c b/drivers/md/bcache/nvmpg.c
> > index a920779eb548..8ce0c4389b42 100644
> > --- a/drivers/md/bcache/nvmpg.c
> > +++ b/drivers/md/bcache/nvmpg.c
> > @@ -248,6 +248,57 @@ static int init_nvmpg_set_header(struct
> bch_nvmpg_ns *ns)
> > return rc;
> > }
> >
> > +static void __free_space(struct bch_nvmpg_ns *ns, unsigned long
> nvmpg_offset,
> > + int order)
> > +{
> > + unsigned long add_pages = (1L << order);
> > + pgoff_t pgoff;
> > + struct page *page;
> > + void *va;
> > +
> > + if (nvmpg_offset == 0) {
> > + pr_err("free pages on offset 0\n");
> > + return;
> > + }
> > +
> > + page = bch_nvmpg_va_to_pg(bch_nvmpg_offset_to_ptr(nvmpg_offset));
> > + WARN_ON((!page) || (page->private != order));
> > + pgoff = page->index;
> > +
> > + while (order < BCH_MAX_ORDER - 1) {
> > + struct page *buddy_page;
> > +
> > + pgoff_t buddy_pgoff = pgoff ^ (1L << order);
> > + pgoff_t parent_pgoff = pgoff & ~(1L << order);
> > +
> > + if ((parent_pgoff + (1L << (order + 1)) > ns->pages_total))
> > + break;
> > +
> > + va = bch_nvmpg_pgoff_to_ptr(ns, buddy_pgoff);
> > + buddy_page = bch_nvmpg_va_to_pg(va);
> > + WARN_ON(!buddy_page);
> > +
> > + if (PageBuddy(buddy_page) && (buddy_page->private == order)) {
> > + list_del((struct list_head *)&buddy_page->zone_device_data);
> > + __ClearPageBuddy(buddy_page);
> > + pgoff = parent_pgoff;
> > + order++;
> > + continue;
> > + }
> > + break;
> > + }
> > +
> > + va = bch_nvmpg_pgoff_to_ptr(ns, pgoff);
> > + page = bch_nvmpg_va_to_pg(va);
> > + WARN_ON(!page);
> > + list_add((struct list_head *)&page->zone_device_data,
> > + &ns->free_area[order]);
> > + page->index = pgoff;
> > + set_page_private(page, order);
> > + __SetPageBuddy(page);
> > + ns->free += add_pages;
> > +}
> > +
> > static void bch_nvmpg_init_free_space(struct bch_nvmpg_ns *ns)
> > {
> > unsigned int start, end, pages;
> > @@ -261,21 +312,19 @@ static void bch_nvmpg_init_free_space(struct
> bch_nvmpg_ns *ns)
> > pages = end - start;
> >
> > while (pages) {
> > - void *addr;
> > -
> > for (i = BCH_MAX_ORDER - 1; i >= 0; i--) {
> > if ((pgoff_start % (1L << i) == 0) &&
> > (pages >= (1L << i)))
> > break;
> > }
> >
> > - addr = bch_nvmpg_pgoff_to_ptr(ns, pgoff_start);
> > - page = bch_nvmpg_va_to_pg(addr);
> > + page = bch_nvmpg_va_to_pg(
> > + bch_nvmpg_pgoff_to_ptr(ns, pgoff_start));
> > set_page_private(page, i);
> > page->index = pgoff_start;
> > - __SetPageBuddy(page);
> > - list_add((struct list_head *)&page->zone_device_data,
> > - &ns->free_area[i]);
> > +
> > + /* In order to update ns->free */
> > + __free_space(ns, pgoff_start, i);
>
> Hi,
>
> While testing this patchset, we found that this is probably wrong, pgoff_start
> represent page number here, however, __free_space expect this to be page
> offset. Maybe this should be:
>
> __free_space(ns, pgoff_start << PAGE_SHIFT, i);
>
Yes,you are right. I'll fix in the next version.
Thanks!
Jianpeng.
> Thanks,
> Kuai
next prev parent reply other threads:[~2022-02-22 5:26 UTC|newest]
Thread overview: 28+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-12-12 17:05 [PATCH v13 00/12] bcache for 5.17: enable NVDIMM for bcache journal Coly Li
2021-12-12 17:05 ` [PATCH v13 01/12] bcache: add initial data structures for nvm pages Coly Li
2021-12-12 17:05 ` [PATCH v13 02/12] bcache: initialize the nvm pages allocator Coly Li
2021-12-12 19:34 ` Jens Axboe
2021-12-28 5:29 ` Coly Li
2021-12-12 17:05 ` [PATCH v13 03/12] bcache: initialization of the buddy Coly Li
2021-12-12 20:10 ` Jens Axboe
2021-12-28 5:29 ` Coly Li
2021-12-15 16:20 ` Dan Carpenter
2021-12-28 5:12 ` Coly Li
2021-12-12 17:05 ` [PATCH v13 04/12] bcache: bch_nvmpg_alloc_pages() " Coly Li
2021-12-12 20:14 ` Jens Axboe
2021-12-28 5:29 ` Coly Li
2021-12-12 17:05 ` [PATCH v13 05/12] bcache: bch_nvmpg_free_pages() of the buddy allocator Coly Li
2021-12-12 20:16 ` Jens Axboe
2021-12-28 5:29 ` Coly Li
2022-02-21 12:36 ` yukuai (C)
2022-02-22 5:03 ` Ma, Jianpeng [this message]
2021-12-12 17:05 ` [PATCH v13 06/12] bcache: get recs list head for allocated pages by specific uuid Coly Li
2021-12-12 20:18 ` Jens Axboe
2021-12-12 17:05 ` [PATCH v13 07/12] bcache: use bucket index to set GC_MARK_METADATA for journal buckets in bch_btree_gc_finish() Coly Li
2021-12-12 17:05 ` [PATCH v13 08/12] bcache: add BCH_FEATURE_INCOMPAT_NVDIMM_META into incompat feature set Coly Li
2021-12-12 17:05 ` [PATCH v13 09/12] bcache: initialize bcache journal for NVDIMM meta device Coly Li
2021-12-12 17:05 ` [PATCH v13 10/12] bcache: support storing bcache journal into " Coly Li
2021-12-12 17:05 ` [PATCH v13 11/12] bcache: read jset from NVDIMM pages for journal replay Coly Li
2021-12-12 17:05 ` [PATCH v13 12/12] bcache: add sysfs interface register_nvdimm_meta to register NVDIMM meta device Coly Li
2021-12-12 20:20 ` [PATCH v13 00/12] bcache for 5.17: enable NVDIMM for bcache journal Jens Axboe
2021-12-28 5:29 ` Coly Li
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=BN7PR11MB270659A970DFD7E01F2D4A38FD3B9@BN7PR11MB2706.namprd11.prod.outlook.com \
--to=jianpeng.ma@intel.com \
--cc=axboe@kernel.dk \
--cc=colyli@suse.de \
--cc=dan.j.williams@intel.com \
--cc=hare@suse.de \
--cc=hch@lst.de \
--cc=linux-bcache@vger.kernel.org \
--cc=linux-block@vger.kernel.org \
--cc=qiaowei.ren@intel.com \
--cc=yukuai3@huawei.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).