Skip to contentSkip to conversation

Optimize large folio interaction with deferred split

PATCH Superseded Ryan Roberts 5 participants 21 messages Jul 20, 2023 – Aug 1, 2023

Subsystems (3)

v3 of 9 Superseded Latest is v9 →

lore
Share & tools
lore
#0 lore raw

v3 cover Optimize large folio interaction with deferred split

Hi All,

[Sending v3 to replace yesterday's v2 after Yu Zhou's feedback]

This is v3 of a small series in support of my work to enable the use of large folios for anonymous memory (known as "FLEXIBLE_THP" or "LARGE_ANON_FOLIO") [1]. It first makes it possible to add large, non-pmd-mappable folios to the deferred split queue. Then it modifies zap_pte_range() to batch-remove spans of physically contiguous pages from the rmap, which means that in the common case, we elide the need to ever put the folio on the deferred split queue, thus reducing lock contention and improving performance.

This becomes more visible once we have lots of large anonymous folios in the system, and Huang Ying has suggested solving this needs to be a prerequisit for merging the main FLEXIBLE_THP/LARGE_ANON_FOLIO work.

The series applies on top of v6.5-rc2 and a branch is available at [2].

I don't have a full test run with the latest versions of all the patches on top of the latest baseline, so not posting results formally. I can get these if people feel they are neccessary though. But anecdotally, for the kernel compilation workload, this series reduces kernel time by ~4% and reduces real-time by ~0.4%, compared with [1].

Changes since v2 [3]
--------------------

- patch 2:
- Reworked at Yu Zhou's request to reduce duplicated code.
- page_remove_rmap() now forwards to folio_remove_rmap_range() for the !compound (PMD mapped) case.
- Both page_remove_rmap() and folio_remove_rmap_range() share common epilogue via new helper function __remove_rmap_finish().
- As a result of the changes, I've removed the previous Reviewed-bys.

- other 2 patches are unchanged.

Changes since v1 [4]
--------------------

- patch 2: Modified doc comment for folio_remove_rmap_range()
- patch 2: Hoisted _nr_pages_mapped manipulation out of page loop so its now modified once per folio_remove_rmap_range() call.
- patch 2: Added check that page range is fully contained by folio in folio_remove_rmap_range()
- patch 2: Fixed some nits raised by Huang, Ying for folio_remove_rmap_range()
- patch 3: Support batch-zap of all anon pages, not just those in anon vmas
- patch 3: Renamed various functions to make their use clear
- patch 3: Various minor refactoring/cleanups
- Added Reviewed-By tags - thanks!

[1] https://lore.kernel.org/linux-mm/[email protected]/
[2] https://gitlab.arm.com/linux-arm/linux-rr/-/tree/features/granule_perf/deferredsplit-lkml_v3
[3] https://lore.kernel.org/linux-mm/[email protected]/
[4] https://lore.kernel.org/linux-mm/[email protected]/

Thanks,
Ryan

Ryan Roberts (3):
mm: Allow deferred splitting of arbitrary large anon folios
mm: Implement folio_remove_rmap_range()
mm: Batch-zap large anonymous folio PTE mappings

 include/linux/rmap.h |   2 +
 mm/memory.c          | 120 +++++++++++++++++++++++++++++++++++++++++
 mm/rmap.c            | 125 ++++++++++++++++++++++++++++++++-----------
 3 files changed, 217 insertions(+), 30 deletions(-)

--
2.25.1

#1 lore raw

PATCH v3 1/3 mm: Allow deferred splitting of arbitrary large anon folios

In preparation for the introduction of large folios for anonymous memory, we would like to be able to split them when they have unmapped subpages, in order to free those unused pages under memory pressure. So remove the artificial requirement that the large folio needed to be at least PMD-sized.

---
 mm/rmap.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
mm/rmap.c +2 -2
index 0c0d8857dfce..eb0bb00dae34 100644
@@ -1426,11 +1426,11 @@ void page_remove_rmap(struct page *page, struct vm_area_struct *vma,
 		__lruvec_stat_mod_folio(folio, idx, -nr);
 
 		/*
-		 * Queue anon THP for deferred split if at least one
+		 * Queue anon large folio for deferred split if at least one
 		 * page of the folio is unmapped and at least one page
 		 * is still mapped.
 		 */
-		if (folio_test_pmd_mappable(folio) && folio_test_anon(folio))
+		if (folio_test_large(folio) && folio_test_anon(folio))
 			if (!compound || nr < nr_pmdmapped)
 				deferred_split_folio(folio);
 	}
Signature · 1 line
2.25.1
#2 lore raw

PATCH v3 2/3 mm: Implement folio_remove_rmap_range()

In reply to Ryan Roberts

Like page_remove_rmap() but batch-removes the rmap for a range of pages belonging to a folio. This can provide a small speedup due to less manipuation of the various counters. But more crucially, if removing the rmap for all pages of a folio in a batch, there is no need to (spuriously) add it to the deferred split list, which saves significant cost when there is contention for the split queue lock.

All contained pages are accounted using the order-0 folio (or base page) scheme.

page_remove_rmap() is refactored so that it forwards to folio_remove_rmap_range() for !compound cases, and both functions now share a common epilogue function. The intention here is to avoid duplication of code.

---
 include/linux/rmap.h |   2 +
 mm/rmap.c            | 125 ++++++++++++++++++++++++++++++++-----------
 2 files changed, 97 insertions(+), 30 deletions(-)
include/linux/rmap.h +2 -0
index b87d01660412..f578975c12c0 100644
@@ -200,6 +200,8 @@ void page_add_file_rmap(struct page *, struct vm_area_struct *,
 		bool compound);
 void page_remove_rmap(struct page *, struct vm_area_struct *,
 		bool compound);
+void folio_remove_rmap_range(struct folio *folio, struct page *page,
+		int nr, struct vm_area_struct *vma);
 
 void hugepage_add_anon_rmap(struct page *, struct vm_area_struct *,
 		unsigned long address, rmap_t flags);
mm/rmap.c +95 -30

145 lines not on this page · raw message

Signature · 1 line
2.25.1
#3 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 27 lines · Ryan Roberts

Like page_remove_rmap() but batch-removes the rmap for a range of pages belonging to a folio. This can provide a small speedup due to less manipuation of the various counters. But more crucially, if removing the rmap for all pages of a folio in a batch, there is no need to (spuriously) add it to the deferred split list, which saves significant cost when there is contention for the split queue lock.

All contained pages are accounted using the order-0 folio (or base page) scheme.

page_remove_rmap() is refactored so that it forwards to folio_remove_rmap_range() for !compound cases, and both functions now share a common epilogue function. The intention here is to avoid duplication of code.

---
 include/linux/rmap.h |   2 +
 mm/rmap.c            | 125 ++++++++++++++++++++++++++++++++-----------
 2 files changed, 97 insertions(+), 30 deletions(-)
include/linux/rmap.h +0 -0
index b87d01660412..f578975c12c0 100644

include/linux/rmap.h void page_add_file_rmap(struct page *, struct vm_area_struct *, · line 204

Quoted 5 lines
include/linux/rmap.h +2 -0
@@ -200,6 +200,8 @@ void page_add_file_rmap(struct page *, struct vm_area_struct *,
                bool compound);
 void page_remove_rmap(struct page *, struct vm_area_struct *,
                bool compound);
+void folio_remove_rmap_range(struct folio *folio, struct page *page,
+               int nr, struct vm_area_struct *vma);

I prefer folio_remove_rmap_range(page, nr, vma). Passing both the folio and the starting page seems redundant to me.

Matthew, is there a convention (function names, parameters, etc.) for operations on a range of pages within a folio?

And regarding the refactor, what I have in mind is that folio_remove_rmap_range() is the core API and page_remove_rmap() is just a wrapper around it, i.e., folio_remove_rmap_range(page, 1, vma).

Let me post a diff later and see if it makes sense to you.

#4 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 23 lines · Yu Zhao
Quoted 22 lines · Ryan Roberts

Like page_remove_rmap() but batch-removes the rmap for a range of pages belonging to a folio. This can provide a small speedup due to less manipuation of the various counters. But more crucially, if removing the rmap for all pages of a folio in a batch, there is no need to (spuriously) add it to the deferred split list, which saves significant cost when there is contention for the split queue lock.

All contained pages are accounted using the order-0 folio (or base page) scheme.

page_remove_rmap() is refactored so that it forwards to folio_remove_rmap_range() for !compound cases, and both functions now share a common epilogue function. The intention here is to avoid duplication of code.

---
 include/linux/rmap.h |   2 +
 mm/rmap.c            | 125 ++++++++++++++++++++++++++++++++-----------
 2 files changed, 97 insertions(+), 30 deletions(-)
Quoted 13 lines

include/linux/rmap.h void page_add_file_rmap(struct page *, struct vm_area_struct *, · line 204

Quoted 10 lines
include/linux/rmap.h +2 -0
index b87d01660412..f578975c12c0 100644
@@ -200,6 +200,8 @@ void page_add_file_rmap(struct page *, struct vm_area_struct *,
                bool compound);
 void page_remove_rmap(struct page *, struct vm_area_struct *,
                bool compound);
+void folio_remove_rmap_range(struct folio *folio, struct page *page,
+               int nr, struct vm_area_struct *vma);

I prefer folio_remove_rmap_range(page, nr, vma). Passing both the folio and the starting page seems redundant to me.

I prefer to pass folio explicitly because it makes it clear that all pages in the range must belong to the same folio.

Quoted 7 lines

Matthew, is there a convention (function names, parameters, etc.) for operations on a range of pages within a folio?

And regarding the refactor, what I have in mind is that folio_remove_rmap_range() is the core API and page_remove_rmap() is just a wrapper around it, i.e., folio_remove_rmap_range(page, 1, vma).

I tried to do it that way, but the existing page_remove_rmap() also takes a 'compound' parameter; it can operate on compound, thp pages and uses the alternative accounting scheme in this case.

I could add a compound parameter to folio_remove_rmap_range() but in that case the range parameters don't make sense - when compound is true we are implicitly operating on the whole folio due to the way the accounting is done. So I felt it was clearer for folio_remove_rmap_range() to deal with small page accounting only. page_remove_rmap() forwards to folio_remove_rmap_range() when compound=false and page_remove_rmap() directly deals with the thp accounting when compound=true.

Quoted 2 lines

Let me post a diff later and see if it makes sense to you.

#5 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

In reply to Yu Zhao
Quoted 8 lines · Yu Zhao

include/linux/rmap.h void page_add_file_rmap(struct page *, struct vm_area_struct *, · line 204

Quoted 2 lines
include/linux/rmap.h +2 -0
+void folio_remove_rmap_range(struct folio *folio, struct page *page,
+               int nr, struct vm_area_struct *vma);

I prefer folio_remove_rmap_range(page, nr, vma). Passing both the folio and the starting page seems redundant to me.

Matthew, is there a convention (function names, parameters, etc.) for operations on a range of pages within a folio?

We've been establishing that convention recently, yes. It seems pointless to re-derive the folio from the page when the caller already has the folio. I also like Ryan's point that it reinforces that all pages must be from the same folio.

Quoted 5 lines

And regarding the refactor, what I have in mind is that folio_remove_rmap_range() is the core API and page_remove_rmap() is just a wrapper around it, i.e., folio_remove_rmap_range(page, 1, vma).

Let me post a diff later and see if it makes sense to you.

I think that can make sense. Because we limit to a single page table, specifying 'nr = 1 << PMD_ORDER' is the same as 'compound = true'. Just make it folio, page, nr, vma. I'd actually prefer it as (vma, folio, page, nr), but that isn't the convention we've had in rmap up until now.

#6 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 19 lines · Matthew Wilcox
Quoted 8 lines · Yu Zhao

include/linux/rmap.h void page_add_file_rmap(struct page *, struct vm_area_struct *, · line 204

Quoted 2 lines
include/linux/rmap.h +2 -0
+void folio_remove_rmap_range(struct folio *folio, struct page *page,
+               int nr, struct vm_area_struct *vma);

I prefer folio_remove_rmap_range(page, nr, vma). Passing both the folio and the starting page seems redundant to me.

Matthew, is there a convention (function names, parameters, etc.) for operations on a range of pages within a folio?

We've been establishing that convention recently, yes. It seems pointless to re-derive the folio from the page when the caller already has the folio. I also like Ryan's point that it reinforces that all pages must be from the same folio.

Quoted 4 lines

And regarding the refactor, what I have in mind is that folio_remove_rmap_range() is the core API and page_remove_rmap() is just a wrapper around it, i.e., folio_remove_rmap_range(page, 1, vma).

Quoted 7 lines
Quoted 1 line

Let me post a diff later and see if it makes sense to you.

I think that can make sense. Because we limit to a single page table, specifying 'nr = 1 << PMD_ORDER' is the same as 'compound = true'. Just make it folio, page, nr, vma. I'd actually prefer it as (vma, folio, page, nr), but that isn't the convention we've had in rmap up until now.

IIUC, even if 'nr = 1 << PMD_ORDER', we may remove one PMD 'compound' mapping, or 'nr' PTE mapping. So, we will still need 'compound' (or some better name) as parameter.

--
Best Regards,
Huang, Ying

#7 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 10 lines · Ying
Quoted 5 lines · Matthew Wilcox

I think that can make sense. Because we limit to a single page table, specifying 'nr = 1 << PMD_ORDER' is the same as 'compound = true'. Just make it folio, page, nr, vma. I'd actually prefer it as (vma, folio, page, nr), but that isn't the convention we've had in rmap up until now.

IIUC, even if 'nr = 1 << PMD_ORDER', we may remove one PMD 'compound' mapping, or 'nr' PTE mapping. So, we will still need 'compound' (or some better name) as parameter.

Oh, this is removing ... so you're concerned with the case where we've split the PMD into PTEs, but all the PTEs are still present in a single page table? OK, I don't have a good answer to that. Maybe that torpedoes the whole idea; I'll think about it.

#8 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 8 lines · Matthew Wilcox
Quoted 7 lines · Ying
Quoted 5 lines · Matthew Wilcox

I think that can make sense. Because we limit to a single page table, specifying 'nr = 1 << PMD_ORDER' is the same as 'compound = true'. Just make it folio, page, nr, vma. I'd actually prefer it as (vma, folio, page, nr), but that isn't the convention we've had in rmap up until now.

Quoted 8 lines
Quoted 3 lines

IIUC, even if 'nr = 1 << PMD_ORDER', we may remove one PMD 'compound' mapping, or 'nr' PTE mapping. So, we will still need 'compound' (or some better name) as parameter.

Oh, this is removing ... so you're concerned with the case where we've split the PMD into PTEs, but all the PTEs are still present in a single page table? OK, I don't have a good answer to that. Maybe that torpedoes the whole idea; I'll think about it.

This is exactly why I think the approach I've already taken is the correct one; a 'range' makes no sense when you are dealing with 'compound' pages because you are accounting the entire folio. So surely its better to reflect that by only accounting small pages in the range version of the API.

#9 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 19 lines · Ryan Roberts
Quoted 8 lines · Matthew Wilcox
Quoted 7 lines · Ying
Quoted 5 lines · Matthew Wilcox

I think that can make sense. Because we limit to a single page table, specifying 'nr = 1 << PMD_ORDER' is the same as 'compound = true'. Just make it folio, page, nr, vma. I'd actually prefer it as (vma, folio, page, nr), but that isn't the convention we've had in rmap up until now.

Quoted 8 lines
Quoted 3 lines

IIUC, even if 'nr = 1 << PMD_ORDER', we may remove one PMD 'compound' mapping, or 'nr' PTE mapping. So, we will still need 'compound' (or some better name) as parameter.

Oh, this is removing ... so you're concerned with the case where we've split the PMD into PTEs, but all the PTEs are still present in a single page table? OK, I don't have a good answer to that. Maybe that torpedoes the whole idea; I'll think about it.

Quoted 4 lines

This is exactly why I think the approach I've already taken is the correct one; a 'range' makes no sense when you are dealing with 'compound' pages because you are accounting the entire folio. So surely its better to reflect that by only accounting small pages in the range version of the API.

If the argument is the compound case is a separate one, then why not a separate API for it?

I don't really care about whether we think 'range' makes sense for 'compound' or not. What I'm saying is:
1. if they are considered one general case, then one API with the compound parameter.
2. if they are considered two specific cases, there should be two APIs. This common design pattern is cleaner IMO.

Right now we have an overlap (redundancy) -- people would have to do two code searches: one for page_remove_rmap() and the other for folio_remove_rmap_range(nr=1), and this IMO is a bad design pattern.

#10 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 28 lines · Yu Zhao
Quoted 19 lines · Ryan Roberts
Quoted 8 lines · Matthew Wilcox
Quoted 7 lines · Ying
Quoted 5 lines · Matthew Wilcox

I think that can make sense. Because we limit to a single page table, specifying 'nr = 1 << PMD_ORDER' is the same as 'compound = true'. Just make it folio, page, nr, vma. I'd actually prefer it as (vma, folio, page, nr), but that isn't the convention we've had in rmap up until now.

Quoted 8 lines
Quoted 3 lines

IIUC, even if 'nr = 1 << PMD_ORDER', we may remove one PMD 'compound' mapping, or 'nr' PTE mapping. So, we will still need 'compound' (or some better name) as parameter.

Oh, this is removing ... so you're concerned with the case where we've split the PMD into PTEs, but all the PTEs are still present in a single page table? OK, I don't have a good answer to that. Maybe that torpedoes the whole idea; I'll think about it.

Quoted 4 lines

This is exactly why I think the approach I've already taken is the correct one; a 'range' makes no sense when you are dealing with 'compound' pages because you are accounting the entire folio. So surely its better to reflect that by only accounting small pages in the range version of the API.

If the argument is the compound case is a separate one, then why not a separate API for it?

Quoted 6 lines

I don't really care about whether we think 'range' makes sense for 'compound' or not. What I'm saying is:
1. if they are considered one general case, then one API with the compound parameter.
2. if they are considered two specific cases, there should be two APIs. This common design pattern is cleaner IMO.

Option 2 definitely makes sense to me and I agree that it would be cleaner to have 2 separate APIs, one for small-page accounting (which can accept a range within a folio) and one for large-page accounting (i.e. compound=true in today's API).

But...

1) That's not how the rest of the rmap API does it

2) This would be a much bigger change since I'm removing an existing API and replacing it with a completely new one (there are ~20 call sites to fix up). I was trying to keep the change small and manageable by maintaining the current API but moving all the small-page logic to the new API, so the old API is a wrapper in that case.

3) You would also need an API for the hugetlb case, which page_remove_rmap() handles today. Perhaps that could also be done by the new API that handles the compound case. But then you are mixing and matching your API styles - one caters for 1 specific case, and the other caters for 2 cases and figures out which one.

Quoted 4 lines

Right now we have an overlap (redundancy) -- people would have to do two code searches: one for page_remove_rmap() and the other for folio_remove_rmap_range(nr=1), and this IMO is a bad design pattern.

I'm open to doing the work to remove this redundancy, but I'd like to hear concensus on this thread that its the right approach first. Although personally I don't see a problem with what I've already done; If you want to operate on a page (inc the old concept of a "compound page" and a hugetlb page) call the old one. If you want to operate on a range of pages in a folio, call the new one.

Thanks,
Ryan

#11 lore raw

Re: [PATCH v3 2/3] mm: Implement folio_remove_rmap_range()

Quoted 37 lines · Ryan Roberts
Quoted 28 lines · Yu Zhao
Quoted 19 lines · Ryan Roberts
Quoted 8 lines · Matthew Wilcox
Quoted 7 lines · Ying
Quoted 5 lines · Matthew Wilcox

I think that can make sense. Because we limit to a single page table, specifying 'nr = 1 << PMD_ORDER' is the same as 'compound = true'. Just make it folio, page, nr, vma. I'd actually prefer it as (vma, folio, page, nr), but that isn't the convention we've had in rmap up until now.

Quoted 8 lines
Quoted 3 lines

IIUC, even if 'nr = 1 << PMD_ORDER', we may remove one PMD 'compound' mapping, or 'nr' PTE mapping. So, we will still need 'compound' (or some better name) as parameter.

Oh, this is removing ... so you're concerned with the case where we've split the PMD into PTEs, but all the PTEs are still present in a single page table? OK, I don't have a good answer to that. Maybe that torpedoes the whole idea; I'll think about it.

Quoted 4 lines

This is exactly why I think the approach I've already taken is the correct one; a 'range' makes no sense when you are dealing with 'compound' pages because you are accounting the entire folio. So surely its better to reflect that by only accounting small pages in the range version of the API.

If the argument is the compound case is a separate one, then why not a separate API for it?

Quoted 6 lines

I don't really care about whether we think 'range' makes sense for 'compound' or not. What I'm saying is:
1. if they are considered one general case, then one API with the compound parameter.
2. if they are considered two specific cases, there should be two APIs. This common design pattern is cleaner IMO.

Quoted 8 lines

Option 2 definitely makes sense to me and I agree that it would be cleaner to have 2 separate APIs, one for small-page accounting (which can accept a range within a folio) and one for large-page accounting (i.e. compound=true in today's API).

But...

1) That's not how the rest of the rmap API does it

Yes, but that's how we convert things: one step a time.

Quoted 5 lines

2) This would be a much bigger change since I'm removing an existing API and replacing it with a completely new one (there are ~20 call sites to fix up). I was trying to keep the change small and manageable by maintaining the current API but moving all the small-page logic to the new API, so the old API is a wrapper in that case.

I don't get how it'd be "much bigger". Isn't it just a straightforward replacement?

Quoted 4 lines

3) You would also need an API for the hugetlb case, which page_remove_rmap() handles today. Perhaps that could also be done by the new API that handles the compound case. But then you are mixing and matching your API styles - one caters for 1 specific case, and the other caters for 2 cases and figures out which one.

You are talking about cases *inside* the APIs, and that's irrelevant to the number of APIs: we only need two -- one supports a range within a folio and the other takes a folio as a single unit.

Quoted 9 lines
Quoted 3 lines

Right now we have an overlap (redundancy) -- people would have to do two code searches: one for page_remove_rmap() and the other for folio_remove_rmap_range(nr=1), and this IMO is a bad design pattern.

I'm open to doing the work to remove this redundancy, but I'd like to hear concensus on this thread that its the right approach first. Although personally I don't see a problem with what I've already done; If you want to operate on a page (inc the old concept of a "compound page" and a hugetlb page) call the old one. If you want to operate on a range of pages in a folio, call the new one.

#12 lore raw

PATCH v3 3/3 mm: Batch-zap large anonymous folio PTE mappings

In reply to Ryan Roberts

This allows batching the rmap removal with folio_remove_rmap_range(), which means we avoid spuriously adding a partially unmapped folio to the deferred split queue in the common case, which reduces split queue lock contention.

Previously each page was removed from the rmap individually with page_remove_rmap(). If the first page belonged to a large folio, this would cause page_remove_rmap() to conclude that the folio was now partially mapped and add the folio to the deferred split queue. But subsequent calls would cause the folio to become fully unmapped, meaning there is no value to adding it to the split queue.

---
 mm/memory.c | 120 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 120 insertions(+)
mm/memory.c +120 -0
index 01f39e8144ef..189b1cfd823d 100644
@@ -1391,6 +1391,94 @@ zap_install_uffd_wp_if_needed(struct vm_area_struct *vma,
 	pte_install_uffd_wp_if_needed(vma, addr, pte, pteval);
 }
 
+static inline unsigned long page_cont_mapped_vaddr(struct page *page,
+				struct page *anchor, unsigned long anchor_vaddr)
+{
+	unsigned long offset;
+	unsigned long vaddr;
+
+	offset = (page_to_pfn(page) - page_to_pfn(anchor)) << PAGE_SHIFT;
+	vaddr = anchor_vaddr + offset;
+
+	if (anchor > page) {
+		if (vaddr > anchor_vaddr)
+			return 0;
+	} else {
+		if (vaddr < anchor_vaddr)
+			return ULONG_MAX;
+	}
+
+	return vaddr;
+}
+
+static int folio_nr_pages_cont_mapped(struct folio *folio,
+				      struct page *page, pte_t *pte,
+				      unsigned long addr, unsigned long end)
+{
+	pte_t ptent;
+	int floops;
+	int i;
+	unsigned long pfn;
+	struct page *folio_end;
+
+	if (!folio_test_large(folio))
+		return 1;
+
+	folio_end = &folio->page + folio_nr_pages(folio);
+	end = min(page_cont_mapped_vaddr(folio_end, page, addr), end);
+	floops = (end - addr) >> PAGE_SHIFT;
+	pfn = page_to_pfn(page);
+	pfn++;
+	pte++;
+
+	for (i = 1; i < floops; i++) {
+		ptent = ptep_get(pte);
+
+		if (!pte_present(ptent) || pte_pfn(ptent) != pfn)
+			break;
+
+		pfn++;
+		pte++;
+	}
+
+	return i;
+}
+
+static unsigned long try_zap_anon_pte_range(struct mmu_gather *tlb,
+					    struct vm_area_struct *vma,
+					    struct folio *folio,
+					    struct page *page, pte_t *pte,
+					    unsigned long addr, int nr_pages,
+					    struct zap_details *details)
+{
+	struct mm_struct *mm = tlb->mm;
+	pte_t ptent;
+	bool full;
+	int i;
+
+	for (i = 0; i < nr_pages;) {
+		ptent = ptep_get_and_clear_full(mm, addr, pte, tlb->fullmm);
+		tlb_remove_tlb_entry(tlb, pte, addr);
+		zap_install_uffd_wp_if_needed(vma, addr, pte, details, ptent);
+		full = __tlb_remove_page(tlb, page, 0);
+
+		if (unlikely(page_mapcount(page) < 1))
+			print_bad_pte(vma, addr, ptent, page);
+
+		i++;
+		page++;
+		pte++;
+		addr += PAGE_SIZE;
+
+		if (unlikely(full))
+			break;
+	}
+
+	folio_remove_rmap_range(folio, page - i, i, vma);
+
+	return i;
+}
+
 static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				struct vm_area_struct *vma, pmd_t *pmd,
 				unsigned long addr, unsigned long end,
@@ -1428,6 +1516,38 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			page = vm_normal_page(vma, addr, ptent);
 			if (unlikely(!should_zap_page(details, page)))
 				continue;
+
+			/*
+			 * Batch zap large anonymous folio mappings. This allows
+			 * batching the rmap removal, which means we avoid
+			 * spuriously adding a partially unmapped folio to the
+			 * deferrred split queue in the common case, which
+			 * reduces split queue lock contention.
+			 */
+			if (page && PageAnon(page)) {
+				struct folio *folio = page_folio(page);
+				int nr_pages_req, nr_pages;
+
+				nr_pages_req = folio_nr_pages_cont_mapped(
+						folio, page, pte, addr, end);
+
+				nr_pages = try_zap_anon_pte_range(tlb, vma,
+						folio, page, pte, addr,
+						nr_pages_req, details);
+
+				rss[mm_counter(page)] -= nr_pages;
+				nr_pages--;
+				pte += nr_pages;
+				addr += nr_pages << PAGE_SHIFT;
+
+				if (unlikely(nr_pages < nr_pages_req)) {
+					force_flush = 1;
+					addr += PAGE_SIZE;
+					break;
+				}
+				continue;
+			}
+
 			ptent = ptep_get_and_clear_full(mm, addr, pte,
 							tlb->fullmm);
 			tlb_remove_tlb_entry(tlb, pte, addr);
Signature · 1 line
2.25.1
#13 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

Hi Ryan,

Quoted 158 lines · Ryan Roberts

This allows batching the rmap removal with folio_remove_rmap_range(), which means we avoid spuriously adding a partially unmapped folio to the deferred split queue in the common case, which reduces split queue lock contention.

Previously each page was removed from the rmap individually with page_remove_rmap(). If the first page belonged to a large folio, this would cause page_remove_rmap() to conclude that the folio was now partially mapped and add the folio to the deferred split queue. But subsequent calls would cause the folio to become fully unmapped, meaning there is no value to adding it to the split queue.

---
 mm/memory.c | 120 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 120 insertions(+)
mm/memory.c +120 -0
index 01f39e8144ef..189b1cfd823d 100644
@@ -1391,6 +1391,94 @@ zap_install_uffd_wp_if_needed(struct vm_area_struct *vma,
 	pte_install_uffd_wp_if_needed(vma, addr, pte, pteval);
 }
 
+static inline unsigned long page_cont_mapped_vaddr(struct page *page,
+				struct page *anchor, unsigned long anchor_vaddr)
+{
+	unsigned long offset;
+	unsigned long vaddr;
+
+	offset = (page_to_pfn(page) - page_to_pfn(anchor)) << PAGE_SHIFT;
+	vaddr = anchor_vaddr + offset;
+
+	if (anchor > page) {
+		if (vaddr > anchor_vaddr)
+			return 0;
+	} else {
+		if (vaddr < anchor_vaddr)
+			return ULONG_MAX;
+	}
+
+	return vaddr;
+}
+
+static int folio_nr_pages_cont_mapped(struct folio *folio,
+				      struct page *page, pte_t *pte,
+				      unsigned long addr, unsigned long end)
+{
+	pte_t ptent;
+	int floops;
+	int i;
+	unsigned long pfn;
+	struct page *folio_end;
+
+	if (!folio_test_large(folio))
+		return 1;
+
+	folio_end = &folio->page + folio_nr_pages(folio);
+	end = min(page_cont_mapped_vaddr(folio_end, page, addr), end);
+	floops = (end - addr) >> PAGE_SHIFT;
+	pfn = page_to_pfn(page);
+	pfn++;
+	pte++;
+
+	for (i = 1; i < floops; i++) {
+		ptent = ptep_get(pte);
+
+		if (!pte_present(ptent) || pte_pfn(ptent) != pfn)
+			break;
+
+		pfn++;
+		pte++;
+	}
+
+	return i;
+}
+
+static unsigned long try_zap_anon_pte_range(struct mmu_gather *tlb,
+					    struct vm_area_struct *vma,
+					    struct folio *folio,
+					    struct page *page, pte_t *pte,
+					    unsigned long addr, int nr_pages,
+					    struct zap_details *details)
+{
+	struct mm_struct *mm = tlb->mm;
+	pte_t ptent;
+	bool full;
+	int i;
+
+	for (i = 0; i < nr_pages;) {
+		ptent = ptep_get_and_clear_full(mm, addr, pte, tlb->fullmm);
+		tlb_remove_tlb_entry(tlb, pte, addr);
+		zap_install_uffd_wp_if_needed(vma, addr, pte, details, ptent);
+		full = __tlb_remove_page(tlb, page, 0);
+
+		if (unlikely(page_mapcount(page) < 1))
+			print_bad_pte(vma, addr, ptent, page);
+
+		i++;
+		page++;
+		pte++;
+		addr += PAGE_SIZE;
+
+		if (unlikely(full))
+			break;
+	}
+
+	folio_remove_rmap_range(folio, page - i, i, vma);
+
+	return i;
+}
+
 static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				struct vm_area_struct *vma, pmd_t *pmd,
 				unsigned long addr, unsigned long end,
@@ -1428,6 +1516,38 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			page = vm_normal_page(vma, addr, ptent);
 			if (unlikely(!should_zap_page(details, page)))
 				continue;
+
+			/*
+			 * Batch zap large anonymous folio mappings. This allows
+			 * batching the rmap removal, which means we avoid
+			 * spuriously adding a partially unmapped folio to the
+			 * deferrred split queue in the common case, which
+			 * reduces split queue lock contention.
+			 */
+			if (page && PageAnon(page)) {
+				struct folio *folio = page_folio(page);
+				int nr_pages_req, nr_pages;
+
+				nr_pages_req = folio_nr_pages_cont_mapped(
+						folio, page, pte, addr, end);
+
+				nr_pages = try_zap_anon_pte_range(tlb, vma,
+						folio, page, pte, addr,
+						nr_pages_req, details);
+
+				rss[mm_counter(page)] -= nr_pages;
+				nr_pages--;
+				pte += nr_pages;
+				addr += nr_pages << PAGE_SHIFT;
+
+				if (unlikely(nr_pages < nr_pages_req)) {
+					force_flush = 1;
+					addr += PAGE_SIZE;
+					break;
+				}
+				continue;
+			}
+
 			ptent = ptep_get_and_clear_full(mm, addr, pte,
 							tlb->fullmm);
 			tlb_remove_tlb_entry(tlb, pte, addr);
Signature · 1 line
2.25.1

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

$ qemu-system-s390x \
    -display none \
    -nodefaults \
    -M s390-ccw-virtio \
    -kernel arch/s390/boot/bzImage \
    -initrd rootfs.cpio \
    -m 512m \
    -serial mon:stdio
KASLR disabled: CPU has no PRNG
KASLR disabled: CPU has no PRNG
[    2.502282] Linux version 6.5.0-rc3+ ([email protected]) (s390-linux-gcc (GCC) 13.1.0, GNU ld (GNU Binutils) 2.40) #1 SMP Wed Jul 26 09:14:20 MST 2023
...
[    3.406011] Freeing initrd memory: 7004K
[    3.492739] BUG: Bad page state in process modprobe  pfn:01b18
[    3.492909] page:00000000233d9f2f refcount:0 mapcount:1 mapping:0000000000000000 index:0xdb pfn:0x1b18
[    3.492998] flags: 0xa0004(uptodate|mappedtodisk|swapbacked|zone=0)
[    3.493195] page_type: 0x0()
[    3.493457] raw: 00000000000a0004 0000000000000100 0000000000000122 0000000000000000
[    3.493492] raw: 00000000000000db 0000000000000000 0000000000000000 0000000000000000
[    3.493525] page dumped because: nonzero mapcount
[    3.493549] Modules linked in:
[    3.493719] CPU: 0 PID: 38 Comm: modprobe Not tainted 6.5.0-rc3+ #1
[    3.493814] Hardware name: QEMU 8561 QEMU (KVM/Linux)
[    3.493892] Call Trace:
[    3.494117]  [<0000000000add35a>] dump_stack_lvl+0x62/0x88
[    3.494333]  [<00000000003d565a>] bad_page+0x8a/0x130
[    3.494355]  [<00000000003d6728>] free_unref_page_prepare+0x268/0x3d8
[    3.494375]  [<00000000003d9408>] free_unref_page+0x48/0x140
[    3.494394]  [<00000000003ad99c>] unmap_page_range+0x924/0x1388
[    3.494412]  [<00000000003ae54c>] unmap_vmas+0x14c/0x200
[    3.494429]  [<00000000003be2f2>] exit_mmap+0xba/0x3a0
[    3.494447]  [<0000000000147000>] __mmput+0x50/0x180
[    3.494466]  [<0000000000152a00>] do_exit+0x320/0xb40
[    3.494484]  [<0000000000153450>] do_group_exit+0x40/0xb8
[    3.494502]  [<00000000001534f6>] __s390x_sys_exit_group+0x2e/0x30
[    3.494520]  [<0000000000b05080>] __do_syscall+0x1e8/0x210
[    3.494539]  [<0000000000b15970>] system_call+0x70/0x98
[    3.494663] Disabling lock debugging due to kernel taint
[    3.494809] BUG: Bad page map in process modprobe  pte:01b1831f pmd:1fff9000
[    3.494833] page:00000000233d9f2f refcount:0 mapcount:0 mapping:0000000000000000 index:0xdb pfn:0x1b18
[    3.494852] flags: 0xa0004(uptodate|mappedtodisk|swapbacked|zone=0)
[    3.494866] page_type: 0xffffffff()
[    3.494882] raw: 00000000000a0004 0000000000000100 0000000000000122 0000000000000000
[    3.494898] raw: 00000000000000db 0000000000000000 ffffffff00000000 0000000000000000
[    3.494908] page dumped because: bad pte
[    3.494923] addr:000002aa1d75c000 vm_flags:08100071 anon_vma:000000001fffc340 mapping:000000000286d6b8 index:db
[    3.494983] file:busybox fault:shmem_fault mmap:shmem_mmap read_folio:0x0
[    3.495247] CPU: 0 PID: 38 Comm: modprobe Tainted: G    B              6.5.0-rc3+ #1
[    3.495267] Hardware name: QEMU 8561 QEMU (KVM/Linux)
[    3.495277] Call Trace:
[    3.495285]  [<0000000000add35a>] dump_stack_lvl+0x62/0x88
[    3.495307]  [<00000000003ab30e>] print_bad_pte+0x176/0x2c8
[    3.495324]  [<00000000003ae098>] unmap_page_range+0x1020/0x1388
[    3.495341]  [<00000000003ae54c>] unmap_vmas+0x14c/0x200
[    3.495357]  [<00000000003be2f2>] exit_mmap+0xba/0x3a0
[    3.495375]  [<0000000000147000>] __mmput+0x50/0x180
[    3.495394]  [<0000000000152a00>] do_exit+0x320/0xb40
[    3.495411]  [<0000000000153450>] do_group_exit+0x40/0xb8
[    3.495429]  [<00000000001534f6>] __s390x_sys_exit_group+0x2e/0x30
[    3.495447]  [<0000000000b05080>] __do_syscall+0x1e8/0x210
[    3.495465]  [<0000000000b15970>] system_call+0x70/0x98
...

The rootfs is available at [2] if it is relevant. I am happy to provide any additional information or test patches as necessary.

Cheers,
Nathan

[1]: https://github.com/nathanchance/llvm-kernel-testing/blob/79aa4ab2edc595979366c427cb49f477c7f31c68/configs/debian/s390x.config
[2]: https://github.com/ClangBuiltLinux/boot-utils/releases/download/20230707-182910/s390-rootfs.cpio.zst

# bad: [0ba5d07205771c50789fd9063950aa75e7f1183f] Add linux-next specific files for 20230726
# good: [18b44bc5a67275641fb26f2c54ba7eef80ac5950] ovl: Always reevaluate the file signature for IMA
git bisect start '0ba5d07205771c50789fd9063950aa75e7f1183f' '18b44bc5a67275641fb26f2c54ba7eef80ac5950'
# bad: [8fe1b33ece8f8fe1377082e839817886cb8c0f81] Merge branch 'main' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net-next.git
git bisect bad 8fe1b33ece8f8fe1377082e839817886cb8c0f81
# bad: [932bd67958459da3dc755b5bea7758e9d951dee5] Merge branch 'ti-next' of git://git.kernel.org/pub/scm/linux/kernel/git/ti/linux.git
git bisect bad 932bd67958459da3dc755b5bea7758e9d951dee5
# bad: [a4abec0a3653fb9dfb3ea6cea2ad1d36f507ca97] Merge branch 'perf-tools-next' of git://git.kernel.org/pub/scm/linux/kernel/git/perf/perf-tools-next.git
git bisect bad a4abec0a3653fb9dfb3ea6cea2ad1d36f507ca97
# bad: [5a52022bde252d090e051077af297dcfeff9fd0d] powerpc/book3s64/radix: add debug message to give more details of vmemmap allocation git bisect bad 5a52022bde252d090e051077af297dcfeff9fd0d
# good: [671115657ee2403d18cb849061d7245687d9fdc5] mm/pgtable: notes on pte_offset_map[_lock]()
git bisect good 671115657ee2403d18cb849061d7245687d9fdc5
# good: [26c3a4fe0eb027ff00ad42168c8732db0c0b40d7] arm64/smmu: use TLBI ASID when invalidating entire range
git bisect good 26c3a4fe0eb027ff00ad42168c8732db0c0b40d7
# bad: [8585d0b53780f11cad8dad37997369949e3d5043] mm: memcg: use rstat for non-hierarchical stats
git bisect bad 8585d0b53780f11cad8dad37997369949e3d5043
# bad: [9abfe35eb187c3f79af5bb07c2f9815a480c4965] mm/compaction: correct comment of candidate pfn in fast_isolate_freepages
git bisect bad 9abfe35eb187c3f79af5bb07c2f9815a480c4965
# bad: [208f64c37a4e22b25b8037776c5713545eaf54fa] selftests: line buffer test program's stdout
git bisect bad 208f64c37a4e22b25b8037776c5713545eaf54fa
# good: [08356142587c28b86817646ff318317b5237fdeb] mmu_notifiers: rename invalidate_range notifier
git bisect good 08356142587c28b86817646ff318317b5237fdeb
# good: [652555287069f2c0bbbfaf262eb41638f5c87550] mm: allow deferred splitting of arbitrary large anon folios
git bisect good 652555287069f2c0bbbfaf262eb41638f5c87550
# bad: [904d9713b3b0e64329b2f6d159966b5c737444ff] mm: batch-zap large anonymous folio PTE mappings
git bisect bad 904d9713b3b0e64329b2f6d159966b5c737444ff
# good: [9a7c14665a566fbc1adc2c35982898abc1546525] mm: implement folio_remove_rmap_range()
git bisect good 9a7c14665a566fbc1adc2c35982898abc1546525

# first bad commit: [904d9713b3b0e64329b2f6d159966b5c737444ff] mm: batch-zap large anonymous folio PTE mappings
#14 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

Quoted 158 lines · Nathan Chancellor

Hi Ryan,

Quoted 155 lines · Ryan Roberts

This allows batching the rmap removal with folio_remove_rmap_range(), which means we avoid spuriously adding a partially unmapped folio to the deferred split queue in the common case, which reduces split queue lock contention.

Previously each page was removed from the rmap individually with page_remove_rmap(). If the first page belonged to a large folio, this would cause page_remove_rmap() to conclude that the folio was now partially mapped and add the folio to the deferred split queue. But subsequent calls would cause the folio to become fully unmapped, meaning there is no value to adding it to the split queue.

---
 mm/memory.c | 120 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 120 insertions(+)
mm/memory.c +120 -0
index 01f39e8144ef..189b1cfd823d 100644
@@ -1391,6 +1391,94 @@ zap_install_uffd_wp_if_needed(struct vm_area_struct *vma,
 	pte_install_uffd_wp_if_needed(vma, addr, pte, pteval);
 }
 
+static inline unsigned long page_cont_mapped_vaddr(struct page *page,
+				struct page *anchor, unsigned long anchor_vaddr)
+{
+	unsigned long offset;
+	unsigned long vaddr;
+
+	offset = (page_to_pfn(page) - page_to_pfn(anchor)) << PAGE_SHIFT;
+	vaddr = anchor_vaddr + offset;
+
+	if (anchor > page) {
+		if (vaddr > anchor_vaddr)
+			return 0;
+	} else {
+		if (vaddr < anchor_vaddr)
+			return ULONG_MAX;
+	}
+
+	return vaddr;
+}
+
+static int folio_nr_pages_cont_mapped(struct folio *folio,
+				      struct page *page, pte_t *pte,
+				      unsigned long addr, unsigned long end)
+{
+	pte_t ptent;
+	int floops;
+	int i;
+	unsigned long pfn;
+	struct page *folio_end;
+
+	if (!folio_test_large(folio))
+		return 1;
+
+	folio_end = &folio->page + folio_nr_pages(folio);
+	end = min(page_cont_mapped_vaddr(folio_end, page, addr), end);
+	floops = (end - addr) >> PAGE_SHIFT;
+	pfn = page_to_pfn(page);
+	pfn++;
+	pte++;
+
+	for (i = 1; i < floops; i++) {
+		ptent = ptep_get(pte);
+
+		if (!pte_present(ptent) || pte_pfn(ptent) != pfn)
+			break;
+
+		pfn++;
+		pte++;
+	}
+
+	return i;
+}
+
+static unsigned long try_zap_anon_pte_range(struct mmu_gather *tlb,
+					    struct vm_area_struct *vma,
+					    struct folio *folio,
+					    struct page *page, pte_t *pte,
+					    unsigned long addr, int nr_pages,
+					    struct zap_details *details)
+{
+	struct mm_struct *mm = tlb->mm;
+	pte_t ptent;
+	bool full;
+	int i;
+
+	for (i = 0; i < nr_pages;) {
+		ptent = ptep_get_and_clear_full(mm, addr, pte, tlb->fullmm);
+		tlb_remove_tlb_entry(tlb, pte, addr);
+		zap_install_uffd_wp_if_needed(vma, addr, pte, details, ptent);
+		full = __tlb_remove_page(tlb, page, 0);
+
+		if (unlikely(page_mapcount(page) < 1))
+			print_bad_pte(vma, addr, ptent, page);
+
+		i++;
+		page++;
+		pte++;
+		addr += PAGE_SIZE;
+
+		if (unlikely(full))
+			break;
+	}
+
+	folio_remove_rmap_range(folio, page - i, i, vma);
+
+	return i;
+}
+
 static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				struct vm_area_struct *vma, pmd_t *pmd,
 				unsigned long addr, unsigned long end,
@@ -1428,6 +1516,38 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			page = vm_normal_page(vma, addr, ptent);
 			if (unlikely(!should_zap_page(details, page)))
 				continue;
+
+			/*
+			 * Batch zap large anonymous folio mappings. This allows
+			 * batching the rmap removal, which means we avoid
+			 * spuriously adding a partially unmapped folio to the
+			 * deferrred split queue in the common case, which
+			 * reduces split queue lock contention.
+			 */
+			if (page && PageAnon(page)) {
+				struct folio *folio = page_folio(page);
+				int nr_pages_req, nr_pages;
+
+				nr_pages_req = folio_nr_pages_cont_mapped(
+						folio, page, pte, addr, end);
+
+				nr_pages = try_zap_anon_pte_range(tlb, vma,
+						folio, page, pte, addr,
+						nr_pages_req, details);
+
+				rss[mm_counter(page)] -= nr_pages;
+				nr_pages--;
+				pte += nr_pages;
+				addr += nr_pages << PAGE_SHIFT;
+
+				if (unlikely(nr_pages < nr_pages_req)) {
+					force_flush = 1;
+					addr += PAGE_SIZE;
+					break;
+				}
+				continue;
+			}
+
 			ptent = ptep_get_and_clear_full(mm, addr, pte,
 							tlb->fullmm);
 			tlb_remove_tlb_entry(tlb, pte, addr);
Quoted 8 lines
Quoted 3 lines
Signature · 1 line
2.25.1

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

Thanks for the bug report and sorry for the inconvenience. I'm going to need a little time to figure out a build environment for s390x and get it reproducing. Hopefully I can come back to you tomorrow with a fix.

Thanks,
Ryan

#15 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

Quoted 168 lines · Ryan Roberts
Quoted 158 lines · Nathan Chancellor

Hi Ryan,

Quoted 155 lines · Ryan Roberts

This allows batching the rmap removal with folio_remove_rmap_range(), which means we avoid spuriously adding a partially unmapped folio to the deferred split queue in the common case, which reduces split queue lock contention.

Previously each page was removed from the rmap individually with page_remove_rmap(). If the first page belonged to a large folio, this would cause page_remove_rmap() to conclude that the folio was now partially mapped and add the folio to the deferred split queue. But subsequent calls would cause the folio to become fully unmapped, meaning there is no value to adding it to the split queue.

---
 mm/memory.c | 120 ++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 120 insertions(+)
mm/memory.c +120 -0
index 01f39e8144ef..189b1cfd823d 100644
@@ -1391,6 +1391,94 @@ zap_install_uffd_wp_if_needed(struct vm_area_struct *vma,
 	pte_install_uffd_wp_if_needed(vma, addr, pte, pteval);
 }
 
+static inline unsigned long page_cont_mapped_vaddr(struct page *page,
+				struct page *anchor, unsigned long anchor_vaddr)
+{
+	unsigned long offset;
+	unsigned long vaddr;
+
+	offset = (page_to_pfn(page) - page_to_pfn(anchor)) << PAGE_SHIFT;
+	vaddr = anchor_vaddr + offset;
+
+	if (anchor > page) {
+		if (vaddr > anchor_vaddr)
+			return 0;
+	} else {
+		if (vaddr < anchor_vaddr)
+			return ULONG_MAX;
+	}
+
+	return vaddr;
+}
+
+static int folio_nr_pages_cont_mapped(struct folio *folio,
+				      struct page *page, pte_t *pte,
+				      unsigned long addr, unsigned long end)
+{
+	pte_t ptent;
+	int floops;
+	int i;
+	unsigned long pfn;
+	struct page *folio_end;
+
+	if (!folio_test_large(folio))
+		return 1;
+
+	folio_end = &folio->page + folio_nr_pages(folio);
+	end = min(page_cont_mapped_vaddr(folio_end, page, addr), end);
+	floops = (end - addr) >> PAGE_SHIFT;
+	pfn = page_to_pfn(page);
+	pfn++;
+	pte++;
+
+	for (i = 1; i < floops; i++) {
+		ptent = ptep_get(pte);
+
+		if (!pte_present(ptent) || pte_pfn(ptent) != pfn)
+			break;
+
+		pfn++;
+		pte++;
+	}
+
+	return i;
+}
+
+static unsigned long try_zap_anon_pte_range(struct mmu_gather *tlb,
+					    struct vm_area_struct *vma,
+					    struct folio *folio,
+					    struct page *page, pte_t *pte,
+					    unsigned long addr, int nr_pages,
+					    struct zap_details *details)
+{
+	struct mm_struct *mm = tlb->mm;
+	pte_t ptent;
+	bool full;
+	int i;
+
+	for (i = 0; i < nr_pages;) {
+		ptent = ptep_get_and_clear_full(mm, addr, pte, tlb->fullmm);
+		tlb_remove_tlb_entry(tlb, pte, addr);
+		zap_install_uffd_wp_if_needed(vma, addr, pte, details, ptent);
+		full = __tlb_remove_page(tlb, page, 0);
+
+		if (unlikely(page_mapcount(page) < 1))
+			print_bad_pte(vma, addr, ptent, page);
+
+		i++;
+		page++;
+		pte++;
+		addr += PAGE_SIZE;
+
+		if (unlikely(full))
+			break;
+	}
+
+	folio_remove_rmap_range(folio, page - i, i, vma);
+
+	return i;
+}
+
 static unsigned long zap_pte_range(struct mmu_gather *tlb,
 				struct vm_area_struct *vma, pmd_t *pmd,
 				unsigned long addr, unsigned long end,
@@ -1428,6 +1516,38 @@ static unsigned long zap_pte_range(struct mmu_gather *tlb,
 			page = vm_normal_page(vma, addr, ptent);
 			if (unlikely(!should_zap_page(details, page)))
 				continue;
+
+			/*
+			 * Batch zap large anonymous folio mappings. This allows
+			 * batching the rmap removal, which means we avoid
+			 * spuriously adding a partially unmapped folio to the
+			 * deferrred split queue in the common case, which
+			 * reduces split queue lock contention.
+			 */
+			if (page && PageAnon(page)) {
+				struct folio *folio = page_folio(page);
+				int nr_pages_req, nr_pages;
+
+				nr_pages_req = folio_nr_pages_cont_mapped(
+						folio, page, pte, addr, end);
+
+				nr_pages = try_zap_anon_pte_range(tlb, vma,
+						folio, page, pte, addr,
+						nr_pages_req, details);
+
+				rss[mm_counter(page)] -= nr_pages;
+				nr_pages--;
+				pte += nr_pages;
+				addr += nr_pages << PAGE_SHIFT;
+
+				if (unlikely(nr_pages < nr_pages_req)) {
+					force_flush = 1;
+					addr += PAGE_SIZE;
+					break;
+				}
+				continue;
+			}
+
 			ptent = ptep_get_and_clear_full(mm, addr, pte,
 							tlb->fullmm);
 			tlb_remove_tlb_entry(tlb, pte, addr);
Quoted 8 lines
Quoted 3 lines
Signature · 1 line
2.25.1

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

Quoted 3 lines

Thanks for the bug report and sorry for the inconvenience. I'm going to need a little time to figure out a build environment for s390x and get it reproducing. Hopefully I can come back to you tomorrow with a fix.

No worries! For what it's worth, if you are not already aware of it, there are GCC toolchains on kernel.org, which is what I use in general and in this particular case:

https://kernel.org/pub/tools/crosstool/

You can just download them to somewhere on your drive then use CROSS_COMPILE=.../bin/s390-linux-gnu-, rather than downloading a bunch of distribution packages.

Cheers,
Nathan

For what it's worth, I just use the GCC toolchains that are on kernel.org:

#16 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

In reply to Nathan Chancellor
Quoted 3 lines · Nathan Chancellor

Hi Ryan,

On Thu, Jul 20, 2023 at 12:29:55PM +0100, Ryan Roberts wrote:

...

Quoted 7 lines
Quoted 1 line

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

Quoted 8 lines
$ qemu-system-s390x \
    -display none \
    -nodefaults \
    -M s390-ccw-virtio \
    -kernel arch/s390/boot/bzImage \
    -initrd rootfs.cpio \
    -m 512m \
    -serial mon:stdio

I'm compiling the kernel for next-20230726 using the s390 cross compiler from kernel.org and the config you linked. Then booting with qemu-system-s390x (tried both distro's 4.2.1 and locally built 8.0.3) and the initrd you provided (tried passing it compressed and uncompressed), but I'm always getting a kernel panic due to not finding a rootfs:

$ qemu-system-s390x \
	-display none \
	-nodefaults \
	-M s390-ccw-virtio \
	-kernel arch/s390/boot/bzImage
	-initrd ../s390-rootfs.cpio.zst \
	-m 512m \
	-serial mon:stdio
KASLR disabled: CPU has no PRNG
KASLR disabled: CPU has no PRNG
Linux version 6.5.0-rc3-next-20230726 (ryarob01@e125769) (s390-linux-gcc (GCC) 13.1.0, GNU ld (GNU Binutils) 2.40) #1 SMP Wed Jul 26 19:56:26 BST 2023
setup: Linux is running under KVM in 64-bit mode
setup: The maximum memory size is 512MB
setup: Relocating AMODE31 section of size 0x00003000
cpu: 1 configured CPUs, 0 standby CPUs
Write protected kernel read-only data: 4036k
Zone ranges:
  DMA      [mem 0x0000000000000000-0x000000007fffffff]
  Normal   empty
Movable zone start for each node
Early memory node ranges
  node   0: [mem 0x0000000000000000-0x000000001fffffff]
Initmem setup node 0 [mem 0x0000000000000000-0x000000001fffffff]
percpu: Embedded 14 pages/cpu s26368 r0 d30976 u57344
Kernel command line: 
random: crng init done
Dentry cache hash table entries: 65536 (order: 7, 524288 bytes, linear)
Inode-cache hash table entries: 32768 (order: 6, 262144 bytes, linear)
Built 1 zonelists, mobility grouping on.  Total pages: 129024
mem auto-init: stack:all(zero), heap alloc:off, heap free:off
Memory: 507720K/524288K available (3464K kernel code, 788K rwdata, 572K rodata, 796K init, 400K bss, 16568K reserved, 0K cma-reserved)
SLUB: HWalign=256, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
rcu: Hierarchical RCU implementation.
rcu:    RCU restricting CPUs from NR_CPUS=64 to nr_cpu_ids=1.
rcu: RCU calculated value of scheduler-enlistment delay is 25 jiffies.
rcu: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=1
NR_IRQS: 3, nr_irqs: 3, preallocated irqs: 3
rcu: srcu_init: Setting srcu_struct sizes based on contention.
clocksource: tod: mask: 0xffffffffffffffff max_cycles: 0x3b0a9be803b0a9, max_idle_ns: 1805497147909793 ns
Console: colour dummy device 80x25
printk: console [ttysclp0] enabled
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
Mountpoint-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
rcu: Hierarchical SRCU implementation.
rcu:    Max phase no-delay instances is 1000.
smp: Bringing up secondary CPUs ...
smp: Brought up 1 node, 1 CPU
clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
futex hash table entries: 256 (order: 4, 65536 bytes, linear)
kvm-s390: SIE is not available
hypfs: The hardware system does not support hypfs
workingset: timestamp_bits=62 max_order=17 bucket_order=0
io scheduler mq-deadline registered
io scheduler kyber registered
cio: Channel measurement facility initialized using format extended (mode autodetected)
Discipline DIAG cannot be used without z/VM
vmur: The z/VM virtual unit record device driver cannot be loaded without z/VM
sclp_sd: Store Data request failed (eq=2, di=3, response=0x40f0, flags=0x00, status=0, rc=-5)
List of all partitions:
No filesystem could mount root, tried: 
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(1,0)
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.5.0-rc3-next-20230726 #1
Hardware name: QEMU 8561 QEMU (KVM/Linux)
Call Trace:
 [<0000000000432b22>] dump_stack_lvl+0x62/0x80 
 [<0000000000158898>] panic+0x2f8/0x310 
 [<00000000005b7a56>] mount_root_generic+0x276/0x3b8 
 [<00000000005b7e46>] prepare_namespace+0x56/0x220 
 [<00000000004593e8>] kernel_init+0x28/0x1c8 
 [<000000000010217e>] __ret_from_fork+0x36/0x50 
 [<000000000046143a>] ret_from_fork+0xa/0x30

Any idea what I'm doing wrong here? Do I need to specify something on the kernel command line? (I've tried root=/dev/ram0, but get the same result).

Thanks,
Ryan

#17 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

Quoted 100 lines · Ryan Roberts
Quoted 3 lines · Nathan Chancellor

Hi Ryan,

On Thu, Jul 20, 2023 at 12:29:55PM +0100, Ryan Roberts wrote:

...

Quoted 7 lines
Quoted 1 line

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

Quoted 8 lines
$ qemu-system-s390x \
    -display none \
    -nodefaults \
    -M s390-ccw-virtio \
    -kernel arch/s390/boot/bzImage \
    -initrd rootfs.cpio \
    -m 512m \
    -serial mon:stdio

I'm compiling the kernel for next-20230726 using the s390 cross compiler from kernel.org and the config you linked. Then booting with qemu-system-s390x (tried both distro's 4.2.1 and locally built 8.0.3) and the initrd you provided (tried passing it compressed and uncompressed), but I'm always getting a kernel panic due to not finding a rootfs:

$ qemu-system-s390x \
	-display none \
	-nodefaults \
	-M s390-ccw-virtio \
	-kernel arch/s390/boot/bzImage
	-initrd ../s390-rootfs.cpio.zst \
	-m 512m \
	-serial mon:stdio
KASLR disabled: CPU has no PRNG
KASLR disabled: CPU has no PRNG
Linux version 6.5.0-rc3-next-20230726 (ryarob01@e125769) (s390-linux-gcc (GCC) 13.1.0, GNU ld (GNU Binutils) 2.40) #1 SMP Wed Jul 26 19:56:26 BST 2023
setup: Linux is running under KVM in 64-bit mode
setup: The maximum memory size is 512MB
setup: Relocating AMODE31 section of size 0x00003000
cpu: 1 configured CPUs, 0 standby CPUs
Write protected kernel read-only data: 4036k
Zone ranges:
  DMA      [mem 0x0000000000000000-0x000000007fffffff]
  Normal   empty
Movable zone start for each node
Early memory node ranges
  node   0: [mem 0x0000000000000000-0x000000001fffffff]
Initmem setup node 0 [mem 0x0000000000000000-0x000000001fffffff]
percpu: Embedded 14 pages/cpu s26368 r0 d30976 u57344
Kernel command line: 
random: crng init done
Dentry cache hash table entries: 65536 (order: 7, 524288 bytes, linear)
Inode-cache hash table entries: 32768 (order: 6, 262144 bytes, linear)
Built 1 zonelists, mobility grouping on.  Total pages: 129024
mem auto-init: stack:all(zero), heap alloc:off, heap free:off
Memory: 507720K/524288K available (3464K kernel code, 788K rwdata, 572K rodata, 796K init, 400K bss, 16568K reserved, 0K cma-reserved)
SLUB: HWalign=256, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
rcu: Hierarchical RCU implementation.
rcu:    RCU restricting CPUs from NR_CPUS=64 to nr_cpu_ids=1.
rcu: RCU calculated value of scheduler-enlistment delay is 25 jiffies.
rcu: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=1
NR_IRQS: 3, nr_irqs: 3, preallocated irqs: 3
rcu: srcu_init: Setting srcu_struct sizes based on contention.
clocksource: tod: mask: 0xffffffffffffffff max_cycles: 0x3b0a9be803b0a9, max_idle_ns: 1805497147909793 ns
Console: colour dummy device 80x25
printk: console [ttysclp0] enabled
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
Mountpoint-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
rcu: Hierarchical SRCU implementation.
rcu:    Max phase no-delay instances is 1000.
smp: Bringing up secondary CPUs ...
smp: Brought up 1 node, 1 CPU
clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
futex hash table entries: 256 (order: 4, 65536 bytes, linear)
kvm-s390: SIE is not available
hypfs: The hardware system does not support hypfs
workingset: timestamp_bits=62 max_order=17 bucket_order=0
io scheduler mq-deadline registered
io scheduler kyber registered
cio: Channel measurement facility initialized using format extended (mode autodetected)
Discipline DIAG cannot be used without z/VM
vmur: The z/VM virtual unit record device driver cannot be loaded without z/VM
sclp_sd: Store Data request failed (eq=2, di=3, response=0x40f0, flags=0x00, status=0, rc=-5)
List of all partitions:
No filesystem could mount root, tried: 
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(1,0)
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.5.0-rc3-next-20230726 #1
Hardware name: QEMU 8561 QEMU (KVM/Linux)
Call Trace:
 [<0000000000432b22>] dump_stack_lvl+0x62/0x80 
 [<0000000000158898>] panic+0x2f8/0x310 
 [<00000000005b7a56>] mount_root_generic+0x276/0x3b8 
 [<00000000005b7e46>] prepare_namespace+0x56/0x220 
 [<00000000004593e8>] kernel_init+0x28/0x1c8 
 [<000000000010217e>] __ret_from_fork+0x36/0x50 
 [<000000000046143a>] ret_from_fork+0xa/0x30
Quoted 1 line

Any idea what I'm doing wrong here? Do I need to specify something on the kernel command line? (I've tried root=/dev/ram0, but get the same result).

Hmmm, interesting. The rootfs does need to be used decompressed (I think the kernel does support zstd compressed initrds but we only compress them to save space, not for running). Does the sha256sum sum match the one I just tested?

$ curl -LSsO https://github.com/ClangBuiltLinux/boot-utils/releases/download/20230707-182910/s390-rootfs.cpio.zst
$ zstd -d s390-rootfs.cpio.zst
$ sha256sum s390-rootfs.cpio
948fb3c2ad65e26aee8eb0a069f5c9e1ab2c59e4b4f62b63ead271e12a8479b4  s390-rootfs.cpio
$ qemu-system-s390x -display none -nodefaults -M s390-ccw-virtio -kernel arch/s390/boot/bzImage -initrd s390-rootfs.cpio -m 512m -serial mon:stdio
...
[    7.890385] Trying to unpack rootfs image as initramfs...
...

I suppose it could be something with Kconfig too, here is the actual one that olddefconfig produces for me:

https://gist.github.com/nathanchance/3e4c1721ac204bbb969e2f288e1695c9

Cheers,
Nathan

#18 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

Quoted 108 lines · Nathan Chancellor
Quoted 100 lines · Ryan Roberts
Quoted 3 lines · Nathan Chancellor

Hi Ryan,

On Thu, Jul 20, 2023 at 12:29:55PM +0100, Ryan Roberts wrote:

...

Quoted 7 lines
Quoted 1 line

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

Quoted 8 lines
$ qemu-system-s390x \
    -display none \
    -nodefaults \
    -M s390-ccw-virtio \
    -kernel arch/s390/boot/bzImage \
    -initrd rootfs.cpio \
    -m 512m \
    -serial mon:stdio

I'm compiling the kernel for next-20230726 using the s390 cross compiler from kernel.org and the config you linked. Then booting with qemu-system-s390x (tried both distro's 4.2.1 and locally built 8.0.3) and the initrd you provided (tried passing it compressed and uncompressed), but I'm always getting a kernel panic due to not finding a rootfs:

$ qemu-system-s390x \
	-display none \
	-nodefaults \
	-M s390-ccw-virtio \
	-kernel arch/s390/boot/bzImage
	-initrd ../s390-rootfs.cpio.zst \
	-m 512m \
	-serial mon:stdio
KASLR disabled: CPU has no PRNG
KASLR disabled: CPU has no PRNG
Linux version 6.5.0-rc3-next-20230726 (ryarob01@e125769) (s390-linux-gcc (GCC) 13.1.0, GNU ld (GNU Binutils) 2.40) #1 SMP Wed Jul 26 19:56:26 BST 2023
setup: Linux is running under KVM in 64-bit mode
setup: The maximum memory size is 512MB
setup: Relocating AMODE31 section of size 0x00003000
cpu: 1 configured CPUs, 0 standby CPUs
Write protected kernel read-only data: 4036k
Zone ranges:
  DMA      [mem 0x0000000000000000-0x000000007fffffff]
  Normal   empty
Movable zone start for each node
Early memory node ranges
  node   0: [mem 0x0000000000000000-0x000000001fffffff]
Initmem setup node 0 [mem 0x0000000000000000-0x000000001fffffff]
percpu: Embedded 14 pages/cpu s26368 r0 d30976 u57344
Kernel command line: 
random: crng init done
Dentry cache hash table entries: 65536 (order: 7, 524288 bytes, linear)
Inode-cache hash table entries: 32768 (order: 6, 262144 bytes, linear)
Built 1 zonelists, mobility grouping on.  Total pages: 129024
mem auto-init: stack:all(zero), heap alloc:off, heap free:off
Memory: 507720K/524288K available (3464K kernel code, 788K rwdata, 572K rodata, 796K init, 400K bss, 16568K reserved, 0K cma-reserved)
SLUB: HWalign=256, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
rcu: Hierarchical RCU implementation.
rcu:    RCU restricting CPUs from NR_CPUS=64 to nr_cpu_ids=1.
rcu: RCU calculated value of scheduler-enlistment delay is 25 jiffies.
rcu: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=1
NR_IRQS: 3, nr_irqs: 3, preallocated irqs: 3
rcu: srcu_init: Setting srcu_struct sizes based on contention.
clocksource: tod: mask: 0xffffffffffffffff max_cycles: 0x3b0a9be803b0a9, max_idle_ns: 1805497147909793 ns
Console: colour dummy device 80x25
printk: console [ttysclp0] enabled
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
Mountpoint-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
rcu: Hierarchical SRCU implementation.
rcu:    Max phase no-delay instances is 1000.
smp: Bringing up secondary CPUs ...
smp: Brought up 1 node, 1 CPU
clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
futex hash table entries: 256 (order: 4, 65536 bytes, linear)
kvm-s390: SIE is not available
hypfs: The hardware system does not support hypfs
workingset: timestamp_bits=62 max_order=17 bucket_order=0
io scheduler mq-deadline registered
io scheduler kyber registered
cio: Channel measurement facility initialized using format extended (mode autodetected)
Discipline DIAG cannot be used without z/VM
vmur: The z/VM virtual unit record device driver cannot be loaded without z/VM
sclp_sd: Store Data request failed (eq=2, di=3, response=0x40f0, flags=0x00, status=0, rc=-5)
List of all partitions:
No filesystem could mount root, tried: 
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(1,0)
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.5.0-rc3-next-20230726 #1
Hardware name: QEMU 8561 QEMU (KVM/Linux)
Call Trace:
 [<0000000000432b22>] dump_stack_lvl+0x62/0x80 
 [<0000000000158898>] panic+0x2f8/0x310 
 [<00000000005b7a56>] mount_root_generic+0x276/0x3b8 
 [<00000000005b7e46>] prepare_namespace+0x56/0x220 
 [<00000000004593e8>] kernel_init+0x28/0x1c8 
 [<000000000010217e>] __ret_from_fork+0x36/0x50 
 [<000000000046143a>] ret_from_fork+0xa/0x30
Quoted 1 line

Any idea what I'm doing wrong here? Do I need to specify something on the kernel command line? (I've tried root=/dev/ram0, but get the same result).

Hmmm, interesting. The rootfs does need to be used decompressed (I think the kernel does support zstd compressed initrds but we only compress them to save space, not for running). Does the sha256sum sum match the one I just tested?

Quoted 4 lines
$ curl -LSsO https://github.com/ClangBuiltLinux/boot-utils/releases/download/20230707-182910/s390-rootfs.cpio.zst
$ zstd -d s390-rootfs.cpio.zst
$ sha256sum s390-rootfs.cpio
948fb3c2ad65e26aee8eb0a069f5c9e1ab2c59e4b4f62b63ead271e12a8479b4  s390-rootfs.cpio

Yes, this matches.

Quoted 9 lines
$ qemu-system-s390x -display none -nodefaults -M s390-ccw-virtio -kernel arch/s390/boot/bzImage -initrd s390-rootfs.cpio -m 512m -serial mon:stdio
...
[    7.890385] Trying to unpack rootfs image as initramfs...
...

I suppose it could be something with Kconfig too, here is the actual one that olddefconfig produces for me:

https://gist.github.com/nathanchance/3e4c1721ac204bbb969e2f288e1695c9

Ahh, I think this was the problem, after downloading this, the kernel is taking much longer to compile and eventually giving me an assembler error:

arch/s390/kernel/mcount.S: Assembler messages:
arch/s390/kernel/mcount.S:140: Error: Unrecognized opcode: `aghik' make[4]: *** [scripts/Makefile.build:360: arch/s390/kernel/mcount.o] Error 1 make[4]: *** Waiting for unfinished jobs....
make[3]: *** [scripts/Makefile.build:480: arch/s390/kernel] Error 2 make[2]: *** [scripts/Makefile.build:480: arch/s390] Error 2 make[2]: *** Waiting for unfinished jobs....
make[1]: *** [/data_nvme0n1/ryarob01/granule_perf/linux/Makefile:2035: .] Error 2
make: *** [Makefile:234: __sub-make] Error 2

The assembler that comes with the kernel.org toolchain is from binutils 2.40. Perhaps that's too old for this instruction? It's inside an "#ifdef CONFIG_FUNCTION_GRAPH_TRACER" block, so I'm guessing whatever config I was previously using didn't have that enabled. I'll try to disable some configs to workaround. What assembler are you using?

Quoted 3 lines

Cheers,
Nathan

#19 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

Quoted 139 lines · Ryan Roberts
Quoted 108 lines · Nathan Chancellor
Quoted 100 lines · Ryan Roberts
Quoted 3 lines · Nathan Chancellor

Hi Ryan,

On Thu, Jul 20, 2023 at 12:29:55PM +0100, Ryan Roberts wrote:

...

Quoted 7 lines
Quoted 1 line

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

Quoted 8 lines
$ qemu-system-s390x \
    -display none \
    -nodefaults \
    -M s390-ccw-virtio \
    -kernel arch/s390/boot/bzImage \
    -initrd rootfs.cpio \
    -m 512m \
    -serial mon:stdio

I'm compiling the kernel for next-20230726 using the s390 cross compiler from kernel.org and the config you linked. Then booting with qemu-system-s390x (tried both distro's 4.2.1 and locally built 8.0.3) and the initrd you provided (tried passing it compressed and uncompressed), but I'm always getting a kernel panic due to not finding a rootfs:

$ qemu-system-s390x \
	-display none \
	-nodefaults \
	-M s390-ccw-virtio \
	-kernel arch/s390/boot/bzImage
	-initrd ../s390-rootfs.cpio.zst \
	-m 512m \
	-serial mon:stdio
KASLR disabled: CPU has no PRNG
KASLR disabled: CPU has no PRNG
Linux version 6.5.0-rc3-next-20230726 (ryarob01@e125769) (s390-linux-gcc (GCC) 13.1.0, GNU ld (GNU Binutils) 2.40) #1 SMP Wed Jul 26 19:56:26 BST 2023
setup: Linux is running under KVM in 64-bit mode
setup: The maximum memory size is 512MB
setup: Relocating AMODE31 section of size 0x00003000
cpu: 1 configured CPUs, 0 standby CPUs
Write protected kernel read-only data: 4036k
Zone ranges:
  DMA      [mem 0x0000000000000000-0x000000007fffffff]
  Normal   empty
Movable zone start for each node
Early memory node ranges
  node   0: [mem 0x0000000000000000-0x000000001fffffff]
Initmem setup node 0 [mem 0x0000000000000000-0x000000001fffffff]
percpu: Embedded 14 pages/cpu s26368 r0 d30976 u57344
Kernel command line: 
random: crng init done
Dentry cache hash table entries: 65536 (order: 7, 524288 bytes, linear)
Inode-cache hash table entries: 32768 (order: 6, 262144 bytes, linear)
Built 1 zonelists, mobility grouping on.  Total pages: 129024
mem auto-init: stack:all(zero), heap alloc:off, heap free:off
Memory: 507720K/524288K available (3464K kernel code, 788K rwdata, 572K rodata, 796K init, 400K bss, 16568K reserved, 0K cma-reserved)
SLUB: HWalign=256, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
rcu: Hierarchical RCU implementation.
rcu:    RCU restricting CPUs from NR_CPUS=64 to nr_cpu_ids=1.
rcu: RCU calculated value of scheduler-enlistment delay is 25 jiffies.
rcu: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=1
NR_IRQS: 3, nr_irqs: 3, preallocated irqs: 3
rcu: srcu_init: Setting srcu_struct sizes based on contention.
clocksource: tod: mask: 0xffffffffffffffff max_cycles: 0x3b0a9be803b0a9, max_idle_ns: 1805497147909793 ns
Console: colour dummy device 80x25
printk: console [ttysclp0] enabled
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
Mountpoint-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
rcu: Hierarchical SRCU implementation.
rcu:    Max phase no-delay instances is 1000.
smp: Bringing up secondary CPUs ...
smp: Brought up 1 node, 1 CPU
clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
futex hash table entries: 256 (order: 4, 65536 bytes, linear)
kvm-s390: SIE is not available
hypfs: The hardware system does not support hypfs
workingset: timestamp_bits=62 max_order=17 bucket_order=0
io scheduler mq-deadline registered
io scheduler kyber registered
cio: Channel measurement facility initialized using format extended (mode autodetected)
Discipline DIAG cannot be used without z/VM
vmur: The z/VM virtual unit record device driver cannot be loaded without z/VM
sclp_sd: Store Data request failed (eq=2, di=3, response=0x40f0, flags=0x00, status=0, rc=-5)
List of all partitions:
No filesystem could mount root, tried: 
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(1,0)
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.5.0-rc3-next-20230726 #1
Hardware name: QEMU 8561 QEMU (KVM/Linux)
Call Trace:
 [<0000000000432b22>] dump_stack_lvl+0x62/0x80 
 [<0000000000158898>] panic+0x2f8/0x310 
 [<00000000005b7a56>] mount_root_generic+0x276/0x3b8 
 [<00000000005b7e46>] prepare_namespace+0x56/0x220 
 [<00000000004593e8>] kernel_init+0x28/0x1c8 
 [<000000000010217e>] __ret_from_fork+0x36/0x50 
 [<000000000046143a>] ret_from_fork+0xa/0x30
Quoted 1 line

Any idea what I'm doing wrong here? Do I need to specify something on the kernel command line? (I've tried root=/dev/ram0, but get the same result).

Hmmm, interesting. The rootfs does need to be used decompressed (I think the kernel does support zstd compressed initrds but we only compress them to save space, not for running). Does the sha256sum sum match the one I just tested?

Quoted 4 lines
$ curl -LSsO https://github.com/ClangBuiltLinux/boot-utils/releases/download/20230707-182910/s390-rootfs.cpio.zst
$ zstd -d s390-rootfs.cpio.zst
$ sha256sum s390-rootfs.cpio
948fb3c2ad65e26aee8eb0a069f5c9e1ab2c59e4b4f62b63ead271e12a8479b4  s390-rootfs.cpio

Yes, this matches.

Quoted 9 lines
$ qemu-system-s390x -display none -nodefaults -M s390-ccw-virtio -kernel arch/s390/boot/bzImage -initrd s390-rootfs.cpio -m 512m -serial mon:stdio
...
[    7.890385] Trying to unpack rootfs image as initramfs...
...

I suppose it could be something with Kconfig too, here is the actual one that olddefconfig produces for me:

https://gist.github.com/nathanchance/3e4c1721ac204bbb969e2f288e1695c9

Ahh, I think this was the problem, after downloading this, the kernel is taking much longer to compile and eventually giving me an assembler error:

arch/s390/kernel/mcount.S: Assembler messages:
arch/s390/kernel/mcount.S:140: Error: Unrecognized opcode: `aghik' make[4]: *** [scripts/Makefile.build:360: arch/s390/kernel/mcount.o] Error 1 make[4]: *** Waiting for unfinished jobs....
make[3]: *** [scripts/Makefile.build:480: arch/s390/kernel] Error 2 make[2]: *** [scripts/Makefile.build:480: arch/s390] Error 2 make[2]: *** Waiting for unfinished jobs....
make[1]: *** [/data_nvme0n1/ryarob01/granule_perf/linux/Makefile:2035: .] Error 2
make: *** [Makefile:234: __sub-make] Error 2

Quoted 5 lines

The assembler that comes with the kernel.org toolchain is from binutils 2.40. Perhaps that's too old for this instruction? It's inside an "#ifdef CONFIG_FUNCTION_GRAPH_TRACER" block, so I'm guessing whatever config I was previously using didn't have that enabled. I'll try to disable some configs to workaround. What assembler are you using?

Ah sorry about that, I forgot about that issue because I handled it during my bisect and all the manual reproduction/verification I have been doing in this thread has been against Andrew's -mm tree, which does not have that problem. Apply this patch on top of next-20230726 and everything should work...

https://lore.kernel.org/[email protected]/

Cheers,
Nathan

#20 lore raw

Re: [PATCH v3 3/3] mm: Batch-zap large anonymous folio PTE mappings

Quoted 146 lines · Nathan Chancellor
Quoted 139 lines · Ryan Roberts
Quoted 108 lines · Nathan Chancellor
Quoted 100 lines · Ryan Roberts
Quoted 3 lines · Nathan Chancellor

Hi Ryan,

On Thu, Jul 20, 2023 at 12:29:55PM +0100, Ryan Roberts wrote:

...

Quoted 7 lines
Quoted 1 line

After this change in -next as commit 904d9713b3b0 ("mm: batch-zap large anonymous folio PTE mappings"), I see the following splats several times when booting Debian's s390x configuration (which I have mirrored at [1]) in QEMU (bisect log below):

Quoted 8 lines
$ qemu-system-s390x \
    -display none \
    -nodefaults \
    -M s390-ccw-virtio \
    -kernel arch/s390/boot/bzImage \
    -initrd rootfs.cpio \
    -m 512m \
    -serial mon:stdio

I'm compiling the kernel for next-20230726 using the s390 cross compiler from kernel.org and the config you linked. Then booting with qemu-system-s390x (tried both distro's 4.2.1 and locally built 8.0.3) and the initrd you provided (tried passing it compressed and uncompressed), but I'm always getting a kernel panic due to not finding a rootfs:

$ qemu-system-s390x \
	-display none \
	-nodefaults \
	-M s390-ccw-virtio \
	-kernel arch/s390/boot/bzImage
	-initrd ../s390-rootfs.cpio.zst \
	-m 512m \
	-serial mon:stdio
KASLR disabled: CPU has no PRNG
KASLR disabled: CPU has no PRNG
Linux version 6.5.0-rc3-next-20230726 (ryarob01@e125769) (s390-linux-gcc (GCC) 13.1.0, GNU ld (GNU Binutils) 2.40) #1 SMP Wed Jul 26 19:56:26 BST 2023
setup: Linux is running under KVM in 64-bit mode
setup: The maximum memory size is 512MB
setup: Relocating AMODE31 section of size 0x00003000
cpu: 1 configured CPUs, 0 standby CPUs
Write protected kernel read-only data: 4036k
Zone ranges:
  DMA      [mem 0x0000000000000000-0x000000007fffffff]
  Normal   empty
Movable zone start for each node
Early memory node ranges
  node   0: [mem 0x0000000000000000-0x000000001fffffff]
Initmem setup node 0 [mem 0x0000000000000000-0x000000001fffffff]
percpu: Embedded 14 pages/cpu s26368 r0 d30976 u57344
Kernel command line: 
random: crng init done
Dentry cache hash table entries: 65536 (order: 7, 524288 bytes, linear)
Inode-cache hash table entries: 32768 (order: 6, 262144 bytes, linear)
Built 1 zonelists, mobility grouping on.  Total pages: 129024
mem auto-init: stack:all(zero), heap alloc:off, heap free:off
Memory: 507720K/524288K available (3464K kernel code, 788K rwdata, 572K rodata, 796K init, 400K bss, 16568K reserved, 0K cma-reserved)
SLUB: HWalign=256, Order=0-3, MinObjects=0, CPUs=1, Nodes=1
rcu: Hierarchical RCU implementation.
rcu:    RCU restricting CPUs from NR_CPUS=64 to nr_cpu_ids=1.
rcu: RCU calculated value of scheduler-enlistment delay is 25 jiffies.
rcu: Adjusting geometry for rcu_fanout_leaf=16, nr_cpu_ids=1
NR_IRQS: 3, nr_irqs: 3, preallocated irqs: 3
rcu: srcu_init: Setting srcu_struct sizes based on contention.
clocksource: tod: mask: 0xffffffffffffffff max_cycles: 0x3b0a9be803b0a9, max_idle_ns: 1805497147909793 ns
Console: colour dummy device 80x25
printk: console [ttysclp0] enabled
pid_max: default: 32768 minimum: 301
Mount-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
Mountpoint-cache hash table entries: 1024 (order: 1, 8192 bytes, linear)
rcu: Hierarchical SRCU implementation.
rcu:    Max phase no-delay instances is 1000.
smp: Bringing up secondary CPUs ...
smp: Brought up 1 node, 1 CPU
clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 7645041785100000 ns
futex hash table entries: 256 (order: 4, 65536 bytes, linear)
kvm-s390: SIE is not available
hypfs: The hardware system does not support hypfs
workingset: timestamp_bits=62 max_order=17 bucket_order=0
io scheduler mq-deadline registered
io scheduler kyber registered
cio: Channel measurement facility initialized using format extended (mode autodetected)
Discipline DIAG cannot be used without z/VM
vmur: The z/VM virtual unit record device driver cannot be loaded without z/VM
sclp_sd: Store Data request failed (eq=2, di=3, response=0x40f0, flags=0x00, status=0, rc=-5)
List of all partitions:
No filesystem could mount root, tried: 
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(1,0)
CPU: 0 PID: 1 Comm: swapper/0 Not tainted 6.5.0-rc3-next-20230726 #1
Hardware name: QEMU 8561 QEMU (KVM/Linux)
Call Trace:
 [<0000000000432b22>] dump_stack_lvl+0x62/0x80 
 [<0000000000158898>] panic+0x2f8/0x310 
 [<00000000005b7a56>] mount_root_generic+0x276/0x3b8 
 [<00000000005b7e46>] prepare_namespace+0x56/0x220 
 [<00000000004593e8>] kernel_init+0x28/0x1c8 
 [<000000000010217e>] __ret_from_fork+0x36/0x50 
 [<000000000046143a>] ret_from_fork+0xa/0x30
Quoted 1 line

Any idea what I'm doing wrong here? Do I need to specify something on the kernel command line? (I've tried root=/dev/ram0, but get the same result).

Hmmm, interesting. The rootfs does need to be used decompressed (I think the kernel does support zstd compressed initrds but we only compress them to save space, not for running). Does the sha256sum sum match the one I just tested?

Quoted 4 lines
$ curl -LSsO https://github.com/ClangBuiltLinux/boot-utils/releases/download/20230707-182910/s390-rootfs.cpio.zst
$ zstd -d s390-rootfs.cpio.zst
$ sha256sum s390-rootfs.cpio
948fb3c2ad65e26aee8eb0a069f5c9e1ab2c59e4b4f62b63ead271e12a8479b4  s390-rootfs.cpio

Yes, this matches.

Quoted 9 lines
$ qemu-system-s390x -display none -nodefaults -M s390-ccw-virtio -kernel arch/s390/boot/bzImage -initrd s390-rootfs.cpio -m 512m -serial mon:stdio
...
[    7.890385] Trying to unpack rootfs image as initramfs...
...

I suppose it could be something with Kconfig too, here is the actual one that olddefconfig produces for me:

https://gist.github.com/nathanchance/3e4c1721ac204bbb969e2f288e1695c9

Ahh, I think this was the problem, after downloading this, the kernel is taking much longer to compile and eventually giving me an assembler error:

arch/s390/kernel/mcount.S: Assembler messages:
arch/s390/kernel/mcount.S:140: Error: Unrecognized opcode: `aghik' make[4]: *** [scripts/Makefile.build:360: arch/s390/kernel/mcount.o] Error 1 make[4]: *** Waiting for unfinished jobs....
make[3]: *** [scripts/Makefile.build:480: arch/s390/kernel] Error 2 make[2]: *** [scripts/Makefile.build:480: arch/s390] Error 2 make[2]: *** Waiting for unfinished jobs....
make[1]: *** [/data_nvme0n1/ryarob01/granule_perf/linux/Makefile:2035: .] Error 2
make: *** [Makefile:234: __sub-make] Error 2

Quoted 5 lines

The assembler that comes with the kernel.org toolchain is from binutils 2.40. Perhaps that's too old for this instruction? It's inside an "#ifdef CONFIG_FUNCTION_GRAPH_TRACER" block, so I'm guessing whatever config I was previously using didn't have that enabled. I'll try to disable some configs to workaround. What assembler are you using?

Quoted 7 lines

Ah sorry about that, I forgot about that issue because I handled it during my bisect and all the manual reproduction/verification I have been doing in this thread has been against Andrew's -mm tree, which does not have that problem. Apply this patch on top of next-20230726 and everything should work...

https://lore.kernel.org/[email protected]/

No problem - I just got it working by disabling FUNCTION_TRACER, which avoids aghik, and then disabling generation of BTF (which was complaining that pahole wasn't available).

Anyway, I'm up and running now - can repro what you are seeing. Although it's late in the UK now, so will have to look at this tomorrow. Hopefully I can get to the bottom of it in reasonable time.

Quoted 3 lines

Cheers,
Nathan