Re: [RESEND PATCH v7 03/10] mm: thp: Introduce per-size thp sysfs interface
From: John Hubbard
Date: Wed Nov 29 2023 - 14:41:34 EST
1On 11/29/23 03:05, Ryan Roberts wrote:
On 29/11/2023 03:42, John Hubbard wrote:
On 11/22/23 08:29, Ryan Roberts wrote:
...
+As well as PMD-sized THP described above, it is also possible to
+configure the system to allocate "small-sized THP" to back anonymous
Here's one of the places to change to the new name, which lately is
"multi-size THP", or mTHP or m_thp for short. (I've typed "multi-size"
instead of "multi-sized", because the 'd' doesn't add significantly to
the meaning, and if in doubt, shorter is better.
I was thinking of some light changes to the start of this paragraph, something like:
Modern kernels support "multi-size THP" (mTHP), which introduces the ability to
allocate memory in blocks that are bigger than a base page but smaller than
traditional PMD-size (as described above, in increments of a power-of-2 number
of pages. mTHP can back anonymous
Very nice.
...
By default, PMD-sized hugepages have enabled="inherited" and all other
hugepage sizes have enabled="never".
That all sounds good; will update.
I wonder if "inherit" is even better than "inherited" though?
Yes, I think so. "inherit" was actually my first idea, and after
thinking about it, I just made it worse by adding the "ed". haha. :)
...
Khugepaged controls
-------------------
+.. note::
+ khugepaged currently only searches for opportunities to collapse to
+ PMD-sized THP and no attempt is made to collapse to small-sized
+ THP.
+
"small-sized THP" used here too; I propose to change it to "... collapse to
other THP sizes."
OK, looks good.
khugepaged runs usually at low frequency so while one may not want to
invoke defrag algorithms synchronously during the page faults, it
should be worth invoking defrag at least in khugepaged. However it's
@@ -282,10 +321,11 @@ force
Need of application restart
===========================
-The transparent_hugepage/enabled values and tmpfs mount option only affect
-future behavior. So to make them effective you need to restart any
-application that could have been using hugepages. This also applies to the
-regions registered in khugepaged.
+The transparent_hugepage/enabled and
+transparent_hugepage/hugepages-<size>kB/enabled values and tmpfs mount
+option only affect future behavior. So to make them effective you need
+to restart any application that could have been using hugepages. This
+also applies to the regions registered in khugepaged.
Monitoring usage
================
@@ -308,6 +348,10 @@ frequently will incur overhead.
There are a number of counters in ``/proc/vmstat`` that may be used to
monitor how successfully the system is providing huge pages for use.
+.. note::
+ Currently the below counters only record events relating to
+ PMD-sized THP. Events relating to small-sized THP are not included.
Here's another spot to rename to "multi-size THP".
I propose to change it to "Events relating to other THP sizes..."
I'm also going to move this note to just under the "Monitoring Usage" heading
since its current position doesn't make it clear that it includes
"AnonHugePages". I'll also make it clear in the text that mentions AnonHugePages
counter that it "only applies to traditional PMD-sized THP for historical
reasons" and that "AnonHugePages should have been called AnonHugePmdMapped" as
David suggested in the other thread.
OK.
Are we entirely dropping the AnonHugePtePages that was there in v6? I'm
looking for some way to monitor this. I may have missed it because I haven't
gone through all of v7 yet.
...
-"THPeligible" indicates whether the mapping is eligible for allocating THP
-pages as well as the THP is PMD mappable or not - 1 if true, 0 otherwise.
-It just shows the current status.
+"THPeligible" indicates whether the mapping is eligible for allocating
+naturally aligned THP pages of any currently enabled size. 1 if true, 0
+otherwise. It just shows the current status.
"It just shows the current status"...as opposed to what? I'm missing an
educational opportunity here--not sure what this means. :)
I have no idea either - it seems superfluous. But that sentence was there
already. I can remove it if you like?
Yes, let's remove it, please.
...
+/*
+ * Mask of all large folio orders supported for anonymous THP.
+ */
+#define THP_ORDERS_ALL_ANON BIT(PMD_ORDER)
+
+/*
+ * Mask of all large folio orders supported for file THP.
+ */
+#define THP_ORDERS_ALL_FILE (BIT(PMD_ORDER) | BIT(PUD_ORDER))
Is there something about file THP that implies PUD_ORDER? This
deserves an explanatory comment, I think.
Sorry, I'm not really sure what you are asking? Today's kernel supports
PUD-mapping file-backed memory (mostly DAX I believe). I'm not sure what comment
you want me to add, other than "file-backed memory can support PUD-mapping", but
that's self-evident from the define, isn't it?
Well, it's sort of evident, but confusing to me anyway, because it
combines THP and PUD. Which seems to imply that we have PUD-based THPs
supported, but only for file-backed mappings. Which is weird and needs
some explanation, right?
+
+/*
+ * Mask of all large folio orders supported for THP.
+ */
+#define THP_ORDERS_ALL (THP_ORDERS_ALL_ANON | THP_ORDERS_ALL_FILE)
+
#ifdef CONFIG_TRANSPARENT_HUGEPAGE
#define HPAGE_PMD_SHIFT PMD_SHIFT
#define HPAGE_PMD_SIZE ((1UL) << HPAGE_PMD_SHIFT)
@@ -78,13 +93,18 @@ extern struct kobj_attribute shmem_enabled_attr;
extern unsigned long transparent_hugepage_flags;
-#define hugepage_flags_enabled() \
- (transparent_hugepage_flags & \
- ((1<<TRANSPARENT_HUGEPAGE_FLAG) | \
- (1<<TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG)))
-#define hugepage_flags_always() \
- (transparent_hugepage_flags & \
- (1<<TRANSPARENT_HUGEPAGE_FLAG))
Are macros actually required? If not, static inlines would be nicer.
I agree static inlines are nicer. Here I'm removing existing macros though, and
Oh, I see that I replied in the wrong part of the email, since that's the
*removal* part. oops. :)
...
-static inline bool transhuge_vma_suitable(struct vm_area_struct *vma,
- unsigned long addr)
-{
- unsigned long haddr;
-
- /* Don't have to check pgoff for anonymous vma */
- if (!vma_is_anonymous(vma)) {
- if (!IS_ALIGNED((vma->vm_start >> PAGE_SHIFT) - vma->vm_pgoff,
- HPAGE_PMD_NR))
- return false;
+static inline unsigned long transhuge_vma_suitable(struct vm_area_struct *vma,
+ unsigned long addr, unsigned long orders)
Changing this from a bool to a mask of orders is quite significant, and
both the function name and the function-level comment documentation need
to also be adjusted. Here, perhaps one of these names would work:
transhuge_vma_suitable_orders()
This is my preference, but its getting a bit long. How about:
thp_vma_suitable_orders()
I could then call the other one:
thp_vma_allowable_orders()
So we have a clearly related pair?
Oh yes, that works nicely.
transhuge_vma_orders()>
+{
+ int order;
+
+ /*
+ * Iterate over orders, highest to lowest, removing orders that don't
+ * meet alignment requirements from the set. Exit loop at first order
+ * that meets requirements, since all lower orders must also meet
+ * requirements.
Should we add a little note here, to the effect of, "this is unilaterally
taking over a certain amount of allocation-like policy, by deciding how
to search for folios of a certain size"?
Or is this The Only Way To Do It, after all? I know we had some discussion
about it, and intuitively it feels reasonable, but wanted to poke at it
one last time anyway.
This function isn't trying to implement policy, its just filtering out orders
that don't fit and therefore definitely cannot be used. The result is a set of
orders the could be used and its the policy maker's job to decide which one.
Currently that policy is "biggest one that fits && does not overlap other
non-none ptes && folio successfully allocated". That's implemented in the next
patch and would potentially be swapped out in the future for something more
exotic/optimal.
So I don't think we need any extra comments here.
Ack.
...
-bool hugepage_vma_check(struct vm_area_struct *vma, unsigned long vm_flags,
- bool smaps, bool in_pf, bool enforce_sysfs)
+#define hugepage_global_enabled() \
+ (transparent_hugepage_flags & \
+ ((1<<TRANSPARENT_HUGEPAGE_FLAG) | \
+ (1<<TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG)))
+
+#define hugepage_global_always() \
+ (transparent_hugepage_flags & \
+ (1<<TRANSPARENT_HUGEPAGE_FLAG))
+
Here again, I'd like to request that we avoid macros, as I don't think
they are strictly required.
Yeah I agree. I did them this way, because they already existed and I was just
moving them from the header to here. But I'll change to static inline.
+bool hugepage_flags_enabled(void)
{
+ /*
+ * We cover both the anon and the file-backed case here; we must return
+ * true if globally enabled, even when all anon sizes are set to never.
+ * So we don't need to look at huge_anon_orders_global.
+ */
+ return hugepage_global_enabled() ||
+ huge_anon_orders_always ||
+ huge_anon_orders_madvise;
+}
+
+/**
+ * hugepage_vma_check - determine which hugepage orders can be applied to vma
+ * @vma: the vm area to check
+ * @vm_flags: use these vm_flags instead of vma->vm_flags
+ * @smaps: whether answer will be used for smaps file
+ * @in_pf: whether answer will be used by page fault handler
+ * @enforce_sysfs: whether sysfs config should be taken into account
+ * @orders: bitfield of all orders to consider
+ *
+ * Calculates the intersection of the requested hugepage orders and the allowed
+ * hugepage orders for the provided vma. Permitted orders are encoded as a set
+ * bit at the corresponding bit position (bit-2 corresponds to order-2, bit-3
+ * corresponds to order-3, etc). Order-0 is never considered a hugepage order.
+ *
+ * Return: bitfield of orders allowed for hugepage in the vma. 0 if no hugepage
+ * orders are allowed.
+ */
+unsigned long hugepage_vma_check(struct vm_area_struct *vma,
+ unsigned long vm_flags, bool smaps, bool in_pf,
+ bool enforce_sysfs, unsigned long orders)
Here again, a bool return type has been changed to a bitfield. Let's
also update the function name. Maybe one of these:
hugepage_vma_orders()
hugepage_vma_allowable_orders()
thp_vma_allowable_orders()?
Even better, yes.
thanks,
--
John Hubbard
NVIDIA