linux-kernel - Re: [PATCH 3/4] mm/gup: add a range variant of unpin_user_pages_dirty

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <781583d3-b4d4-2cb0-8e6f-0875f4ba4624@oracle.com>
Date:   Thu, 4 Feb 2021 16:30:03 +0000
From:   Joao Martins <joao.m.martins@...cle.com>
To:     John Hubbard <jhubbard@...dia.com>
Cc:     linux-kernel@...r.kernel.org, linux-rdma@...r.kernel.org,
        Andrew Morton <akpm@...ux-foundation.org>,
        Jason Gunthorpe <jgg@...pe.ca>,
        Doug Ledford <dledford@...hat.com>,
        Matthew Wilcox <willy@...radead.org>, linux-mm@...ck.org
Subject: Re: [PATCH 3/4] mm/gup: add a range variant of
 unpin_user_pages_dirty_lock()

On 2/4/21 11:35 AM, Joao Martins wrote:
> On 2/3/21 11:37 PM, John Hubbard wrote:
>> On 2/3/21 2:00 PM, Joao Martins wrote:
>>> -static inline unsigned int count_ntails(struct page **pages, unsigned long npages)
>>> +static inline unsigned int count_ntails(struct page **pages,
>>> +					unsigned long npages, bool range)
>>>   {
>>> -	struct page *head = compound_head(pages[0]);
>>> +	struct page *page = pages[0], *head = compound_head(page);
>>>   	unsigned int ntails;
>>>   
>>> +	if (range)
>>> +		return (!PageCompound(head) || compound_order(head) <= 1) ? 1 :
>>> +		   min_t(unsigned int, (head + compound_nr(head) - page), npages);
>>
>> Here, you clearly should use a separate set of _range routines. Because you're basically
>> creating two different routines here! Keep it simple.
>>
>> Once you're in a separate routine, you might feel more comfortable expanding that to
>> a more readable form, too:
>>
>> 	if (!PageCompound(head) || compound_order(head) <= 1)
>> 		return 1;
>>
>> 	return min_t(unsigned int, (head + compound_nr(head) - page), npages);
>>
> Yes.
> 
> Let me also try instead to put move everything into two sole iterator helper routines,
> compound_next() and compound_next_range(), and thus get rid of this count_ntails(). It
> should also help in removing a compound_head() call which should save cycles.
> 

As mentioned earlier, I got rid of count_ntails and the ugly boolean. Plus addressed the
missing docs -- fwiw, I borrowed unpin_user_pages_dirty_lock() docs and modified a bit.

Partial diff below, hopefully it is looking better now:

diff --git a/mm/gup.c b/mm/gup.c
index 5a3dd235017a..4ef36c8990e3 100644
--- a/mm/gup.c
+++ b/mm/gup.c
@@ -215,6 +215,34 @@ void unpin_user_page(struct page *page)
 }
 EXPORT_SYMBOL(unpin_user_page);

+static inline void range_next(unsigned long i, unsigned long npages,
+                             struct page **list, struct page **head,
+                             unsigned int *ntails)
+{
+       struct page *next, *page;
+       unsigned int nr = 1;
+
+       if (i >= npages)
+               return;
+
+       npages -= i;
+       next = *list + i;
+
+       page = compound_head(next);
+       if (PageCompound(page) && compound_order(page) > 1)
+               nr = min_t(unsigned int,
+                          page + compound_nr(page) - next, npages);
+
+       *head = page;
+       *ntails = nr;
+}
+
+#define for_each_compound_range(__i, __list, __npages, __head, __ntails) \
+       for (__i = 0, \
+            range_next(__i, __npages, __list, &(__head), &(__ntails)); \
+            __i < __npages; __i += __ntails, \
+            range_next(__i, __npages, __list, &(__head), &(__ntails)))
+
 static inline void compound_next(unsigned long i, unsigned long npages,
                                 struct page **list, struct page **head,
                                 unsigned int *ntails)
@@ -306,6 +334,42 @@ void unpin_user_pages_dirty_lock(struct page **pages, unsigned long
npages,
 }
 EXPORT_SYMBOL(unpin_user_pages_dirty_lock);

+/**
+ * unpin_user_page_range_dirty_lock() - release and optionally dirty
+ * gup-pinned page range
+ *
+ * @page:  the starting page of a range maybe marked dirty, and definitely released.
+ * @npages: number of consecutive pages to release.
+ * @make_dirty: whether to mark the pages dirty
+ *
+ * "gup-pinned page range" refers to a range of pages that has had one of the
+ * get_user_pages() variants called on that page.
+ *
+ * For the page ranges defined by [page .. page+npages], make that range (or
+ * its head pages, if a compound page) dirty, if @make_dirty is true, and if the
+ * page range was previously listed as clean.
+ *
+ * set_page_dirty_lock() is used internally. If instead, set_page_dirty() is
+ * required, then the caller should a) verify that this is really correct,
+ * because _lock() is usually required, and b) hand code it:
+ * set_page_dirty_lock(), unpin_user_page().
+ *
+ */
+void unpin_user_page_range_dirty_lock(struct page *page, unsigned long npages,
+                                     bool make_dirty)
+{
+       unsigned long index;
+       struct page *head;
+       unsigned int ntails;
+
+       for_each_compound_range(index, &page, npages, head, ntails) {
+               if (make_dirty && !PageDirty(head))
+                       set_page_dirty_lock(head);
+               put_compound_head(head, ntails, FOLL_PIN);
+       }
+}
+EXPORT_SYMBOL(unpin_user_page_range_dirty_lock);
+