From: Baoquan He <bhe@redhat.com>
To: David Hildenbrand <david@redhat.com>
Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org,
akpm@linux-foundation.org, richardw.yang@linux.intel.com,
osalvador@suse.de, dan.j.williams@intel.com, mhocko@suse.com,
rppt@linux.ibm.com, robin.murphy@arm.com
Subject: Re: [PATCH v2 3/7] mm/sparse.c: introduce a new function clear_subsection_map()
Date: Tue, 3 Mar 2020 16:38:18 +0800 [thread overview]
Message-ID: <20200303083818.GB4433@MiWiFi-R3L-srv> (raw)
In-Reply-To: <53bba557-149c-54b5-c8d7-27f7b6a55eb0@redhat.com>
On 03/03/20 at 09:33am, David Hildenbrand wrote:
> On 03.03.20 09:22, Baoquan He wrote:
> > On 03/02/20 at 04:43pm, David Hildenbrand wrote:
> >> On 01.03.20 06:20, Baoquan He wrote:
> >>> On 02/28/20 at 03:36pm, David Hildenbrand wrote:
> >>>> On 20.02.20 05:33, Baoquan He wrote:
> >>>>> Wrap the codes which clear subsection map of one memory region from
> >>>>> section_deactivate() into clear_subsection_map().
> >>>>>
> >>>>> Signed-off-by: Baoquan He <bhe@redhat.com>
> >>>>> ---
> >>>>> mm/sparse.c | 46 ++++++++++++++++++++++++++++++++++++++--------
> >>>>> 1 file changed, 38 insertions(+), 8 deletions(-)
> >>>>>
> >>>>> diff --git a/mm/sparse.c b/mm/sparse.c
> >>>>> index 977b47acd38d..df857ee9330c 100644
> >>>>> --- a/mm/sparse.c
> >>>>> +++ b/mm/sparse.c
> >>>>> @@ -726,14 +726,25 @@ static void free_map_bootmem(struct page *memmap)
> >>>>> }
> >>>>> #endif /* CONFIG_SPARSEMEM_VMEMMAP */
> >>>>>
> >>>>> -static void section_deactivate(unsigned long pfn, unsigned long nr_pages,
> >>>>> - struct vmem_altmap *altmap)
> >>>>> +/**
> >>>>> + * clear_subsection_map - Clear subsection map of one memory region
> >>>>> + *
> >>>>> + * @pfn - start pfn of the memory range
> >>>>> + * @nr_pages - number of pfns to add in the region
> >>>>> + *
> >>>>> + * This is only intended for hotplug, and clear the related subsection
> >>>>> + * map inside one section.
> >>>>> + *
> >>>>> + * Return:
> >>>>> + * * -EINVAL - Section already deactived.
> >>>>> + * * 0 - Subsection map is emptied.
> >>>>> + * * 1 - Subsection map is not empty.
> >>>>> + */
> >>>>
> >>>> Less verbose please (in my preference: none and simplify return handling)
> >>>>
> >>>>> +static int clear_subsection_map(unsigned long pfn, unsigned long nr_pages)
> >>>>> {
> >>>>> DECLARE_BITMAP(map, SUBSECTIONS_PER_SECTION) = { 0 };
> >>>>> DECLARE_BITMAP(tmp, SUBSECTIONS_PER_SECTION) = { 0 };
> >>>>> struct mem_section *ms = __pfn_to_section(pfn);
> >>>>> - bool section_is_early = early_section(ms);
> >>>>> - struct page *memmap = NULL;
> >>>>> unsigned long *subsection_map = ms->usage
> >>>>> ? &ms->usage->subsection_map[0] : NULL;
> >>>>>
> >>>>> @@ -744,8 +755,28 @@ static void section_deactivate(unsigned long pfn, unsigned long nr_pages,
> >>>>> if (WARN(!subsection_map || !bitmap_equal(tmp, map, SUBSECTIONS_PER_SECTION),
> >>>>> "section already deactivated (%#lx + %ld)\n",
> >>>>> pfn, nr_pages))
> >>>>> - return;
> >>>>> + return -EINVAL;
> >>>>> +
> >>>>> + bitmap_xor(subsection_map, map, subsection_map, SUBSECTIONS_PER_SECTION);
> >>>>>
> >>>>> + if (bitmap_empty(subsection_map, SUBSECTIONS_PER_SECTION))
> >>>>> + return 0;
> >>>>> +
> >>>>
> >>>> Can we please just have a
> >>>>
> >>>> subsection_map_empty() instead and handle that in the caller?
> >>>> (you can then always return true in the !VMEMMAP variant)
> >>>
> >>> I don't follow. Could you be more specific? or pseudo code please?
> >>>
> >>> The old code has to handle below case in which subsection_map has been
> >>> cleared. And I introduce clear_subsection_map() to encapsulate all
> >>> subsection map realted code so that !VMEMMAP won't have to see it any
> >>> more.
> >>>
> >>
> >> Something like this on top would be easier to understand IMHO
> >>
> >>
> >> diff --git a/mm/sparse.c b/mm/sparse.c
> >> index dc79b00ddaaa..be5c80e9cfee 100644
> >> --- a/mm/sparse.c
> >> +++ b/mm/sparse.c
> >> @@ -726,20 +726,6 @@ static void free_map_bootmem(struct page *memmap)
> >> }
> >> #endif /* CONFIG_SPARSEMEM_VMEMMAP */
> >>
> >> -/**
> >> - * clear_subsection_map - Clear subsection map of one memory region
> >> - *
> >> - * @pfn - start pfn of the memory range
> >> - * @nr_pages - number of pfns to add in the region
> >> - *
> >> - * This is only intended for hotplug, and clear the related subsection
> >> - * map inside one section.
> >> - *
> >> - * Return:
> >> - * * -EINVAL - Section already deactived.
> >> - * * 0 - Subsection map is emptied.
> >> - * * 1 - Subsection map is not empty.
> >> - */
> >> static int clear_subsection_map(unsigned long pfn, unsigned long nr_pages)
> >> {
> >> DECLARE_BITMAP(map, SUBSECTIONS_PER_SECTION) = { 0 };
> >> @@ -758,11 +744,12 @@ static int clear_subsection_map(unsigned long pfn, unsigned long nr_pages)
> >> return -EINVAL;
> >>
> >> bitmap_xor(subsection_map, map, subsection_map, SUBSECTIONS_PER_SECTION);
> >> + return 0;
> >> +}
> >>
> >> - if (bitmap_empty(subsection_map, SUBSECTIONS_PER_SECTION))
> >> - return 0;
> >> -
> >> - return 1;
> >> +static bool is_subsection_map_empty(unsigned long pfn, unsigned long nr_pages)
> >> +{
> >> + return bitmap_empty(subsection_map, SUBSECTIONS_PER_SECTION);
> >> }
> >>
> >> static void section_deactivate(unsigned long pfn, unsigned long nr_pages,
> >> @@ -771,11 +758,8 @@ static void section_deactivate(unsigned long pfn, unsigned long nr_pages,
> >> struct mem_section *ms = __pfn_to_section(pfn);
> >> bool section_is_early = early_section(ms);
> >> struct page *memmap = NULL;
> >> - int rc;
> >> -
> >>
> >> - rc = clear_subsection_map(pfn, nr_pages);
> >> - if (IS_ERR_VALUE((unsigned long)rc))
> >> + if (unlikely(clear_subsection_map(pfn, nr_pages)))
> >> return;
> >> /*
> >> * There are 3 cases to handle across two configurations
> >> @@ -794,7 +778,7 @@ static void section_deactivate(unsigned long pfn, unsigned long nr_pages,
> >> *
> >> * For 2/ and 3/ the SPARSEMEM_VMEMMAP={y,n} cases are unified
> >> */
> >> - if (!rc) {
> >> + if (is_subsection_map_empty(pfn, nr_pages)) {
> >> unsigned long section_nr = pfn_to_section_nr(pfn);
> >
> > Tried this way, it's not good in this patch. Since ms->usage might be
> > freed in this place.
> >
> > if (!PageReserved(virt_to_page(ms->usage))) {
> > kfree(ms->usage);
> > ms->usage = NULL;
> > }
>
> So your patch #1 is already broken. Just cache the result in patch #1.
>
> bool empty;
Right, good catch. Will take this way. Thanks.
>
> ...
> empty = bitmap_empty(subsection_map, SUBSECTIONS_PER_SECTION);
> ...
> if (empty) {
> ...
> }
next prev parent reply other threads:[~2020-03-03 8:38 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-20 4:33 [PATCH v2 0/7] mm/hotplug: Only use subsection map in VMEMMAP case Baoquan He
2020-02-20 4:33 ` [PATCH v2 1/7] mm/hotplug: fix hot remove failure in SPARSEMEM|!VMEMMAP case Baoquan He
2020-02-20 6:11 ` Wei Yang
2020-02-20 12:07 ` David Hildenbrand
2020-02-26 12:31 ` David Hildenbrand
2020-02-26 13:07 ` Baoquan He
2020-02-26 18:09 ` Dan Williams
2020-03-03 8:34 ` David Hildenbrand
2020-03-03 8:39 ` Baoquan He
2020-02-20 4:33 ` [PATCH v2 2/7] mm/sparse.c: introduce new function fill_subsection_map() Baoquan He
2020-02-20 6:14 ` Wei Yang
2020-02-28 14:27 ` David Hildenbrand
2020-03-01 4:59 ` Baoquan He
2020-02-20 4:33 ` [PATCH v2 3/7] mm/sparse.c: introduce a new function clear_subsection_map() Baoquan He
2020-02-20 6:15 ` Wei Yang
2020-02-28 14:36 ` David Hildenbrand
2020-03-01 5:20 ` Baoquan He
2020-03-02 15:43 ` David Hildenbrand
2020-03-03 1:53 ` Baoquan He
2020-03-03 8:22 ` Baoquan He
2020-03-03 8:33 ` David Hildenbrand
2020-03-03 8:38 ` Baoquan He [this message]
2020-02-20 4:33 ` [PATCH v2 4/7] mm/sparse.c: only use subsection map in VMEMMAP case Baoquan He
2020-02-20 6:17 ` Wei Yang
2020-02-25 9:57 ` Michal Hocko
2020-02-26 3:53 ` Baoquan He
2020-02-26 9:10 ` Michal Hocko
2020-02-28 7:25 ` Baoquan He
2020-02-20 4:33 ` [PATCH v2 5/7] mm/sparse.c: add code comment about sub-section hotplug Baoquan He
2020-02-20 4:33 ` [PATCH v2 6/7] mm/sparse.c: move subsection_map related codes together Baoquan He
2020-02-20 6:18 ` Wei Yang
2020-02-20 7:04 ` Baoquan He
2020-02-20 7:12 ` Wei Yang
2020-02-20 8:55 ` Baoquan He
2020-02-20 21:52 ` Wei Yang
2020-02-20 4:33 ` [PATCH v2 7/7] mm/sparse.c: Use __get_free_pages() instead in populate_section_memmap() Baoquan He
2020-02-20 10:38 ` [PATCH v2 0/7] mm/hotplug: Only use subsection map in VMEMMAP case Michal Hocko
2020-02-21 14:28 ` Baoquan He
2020-02-25 9:10 ` David Hildenbrand
2020-02-25 10:02 ` Michal Hocko
2020-02-26 3:42 ` Baoquan He
2020-02-26 9:14 ` Michal Hocko
2020-02-26 12:30 ` Baoquan He
2020-02-25 10:03 ` Michal Hocko
2020-02-26 3:44 ` Baoquan He
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=20200303083818.GB4433@MiWiFi-R3L-srv \
--to=bhe@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=dan.j.williams@intel.com \
--cc=david@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@suse.com \
--cc=osalvador@suse.de \
--cc=richardw.yang@linux.intel.com \
--cc=robin.murphy@arm.com \
--cc=rppt@linux.ibm.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