From: "Aneesh Kumar K.V" <aneesh.kumar@linux.ibm.com>
To: Christophe Leroy <christophe.leroy@csgroup.eu>,
linux-mm@kvack.org, akpm@linux-foundation.org
Cc: linuxppc-dev@lists.ozlabs.org,
Anshuman Khandual <anshuman.khandual@arm.com>
Subject: Re: [PATCH v2 07/13] mm/debug_vm_pgtable/set_pte/pmd/pud: Don't use set_*_at to update an existing pte entry
Date: Fri, 21 Aug 2020 12:44:53 +0530 [thread overview]
Message-ID: <3d966519-0a6b-3ccb-fd21-b7f06c8e4df7@linux.ibm.com> (raw)
In-Reply-To: <b21d1dbb-7439-d317-8516-94c80f333e92@csgroup.eu>
On 8/20/20 8:02 PM, Christophe Leroy wrote:
>
>
> Le 19/08/2020 à 15:01, Aneesh Kumar K.V a écrit :
>> set_pte_at() should not be used to set a pte entry at locations that
>> already holds a valid pte entry. Architectures like ppc64 don't do TLB
>> invalidate in set_pte_at() and hence expect it to be used to set
>> locations
>> that are not a valid PTE.
>>
>> Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
>> ---
>> mm/debug_vm_pgtable.c | 35 +++++++++++++++--------------------
>> 1 file changed, 15 insertions(+), 20 deletions(-)
>>
>> diff --git a/mm/debug_vm_pgtable.c b/mm/debug_vm_pgtable.c
>> index 76f4c713e5a3..9c7e2c9cfc76 100644
>> --- a/mm/debug_vm_pgtable.c
>> +++ b/mm/debug_vm_pgtable.c
>> @@ -74,15 +74,18 @@ static void __init pte_advanced_tests(struct
>> mm_struct *mm,
>> {
>> pte_t pte = pfn_pte(pfn, prot);
>> + /*
>> + * Architectures optimize set_pte_at by avoiding TLB flush.
>> + * This requires set_pte_at to be not used to update an
>> + * existing pte entry. Clear pte before we do set_pte_at
>> + */
>> +
>> pr_debug("Validating PTE advanced\n");
>> pte = pfn_pte(pfn, prot);
>> set_pte_at(mm, vaddr, ptep, pte);
>> ptep_set_wrprotect(mm, vaddr, ptep);
>> pte = ptep_get(ptep);
>> WARN_ON(pte_write(pte));
>> -
>> - pte = pfn_pte(pfn, prot);
>> - set_pte_at(mm, vaddr, ptep, pte);
>> ptep_get_and_clear(mm, vaddr, ptep);
>> pte = ptep_get(ptep);
>> WARN_ON(!pte_none(pte));
>> @@ -96,13 +99,11 @@ static void __init pte_advanced_tests(struct
>> mm_struct *mm,
>> ptep_set_access_flags(vma, vaddr, ptep, pte, 1);
>> pte = ptep_get(ptep);
>> WARN_ON(!(pte_write(pte) && pte_dirty(pte)));
>> -
>> - pte = pfn_pte(pfn, prot);
>> - set_pte_at(mm, vaddr, ptep, pte);
>> ptep_get_and_clear_full(mm, vaddr, ptep, 1);
>> pte = ptep_get(ptep);
>> WARN_ON(!pte_none(pte));
>> + pte = pfn_pte(pfn, prot);
>> pte = pte_mkyoung(pte);
>> set_pte_at(mm, vaddr, ptep, pte);
>> ptep_test_and_clear_young(vma, vaddr, ptep);
>> @@ -164,9 +165,6 @@ static void __init pmd_advanced_tests(struct
>> mm_struct *mm,
>> pmdp_set_wrprotect(mm, vaddr, pmdp);
>> pmd = READ_ONCE(*pmdp);
>> WARN_ON(pmd_write(pmd));
>> -
>> - pmd = pmd_mkhuge(pfn_pmd(pfn, prot));
>> - set_pmd_at(mm, vaddr, pmdp, pmd);
>> pmdp_huge_get_and_clear(mm, vaddr, pmdp);
>> pmd = READ_ONCE(*pmdp);
>> WARN_ON(!pmd_none(pmd));
>> @@ -180,13 +178,11 @@ static void __init pmd_advanced_tests(struct
>> mm_struct *mm,
>> pmdp_set_access_flags(vma, vaddr, pmdp, pmd, 1);
>> pmd = READ_ONCE(*pmdp);
>> WARN_ON(!(pmd_write(pmd) && pmd_dirty(pmd)));
>> -
>> - pmd = pmd_mkhuge(pfn_pmd(pfn, prot));
>> - set_pmd_at(mm, vaddr, pmdp, pmd);
>> pmdp_huge_get_and_clear_full(vma, vaddr, pmdp, 1);
>> pmd = READ_ONCE(*pmdp);
>> WARN_ON(!pmd_none(pmd));
>> + pmd = pmd_mkhuge(pfn_pmd(pfn, prot));
>> pmd = pmd_mkyoung(pmd);
>> set_pmd_at(mm, vaddr, pmdp, pmd);
>> pmdp_test_and_clear_young(vma, vaddr, pmdp);
>> @@ -283,18 +279,10 @@ static void __init pud_advanced_tests(struct
>> mm_struct *mm,
>> WARN_ON(pud_write(pud));
>> #ifndef __PAGETABLE_PMD_FOLDED
>
> Same as below, once set_put_at() is gone, I don't think this #ifndef
> __PAGETABLE_PMD_FOLDED is still need, should be possible to replace by
> 'if (mm_pmd_folded())'
I would skip that change in this series because I still haven't worked
out what it means to have FOLDED PMD with
CONFIG_HAVE_ARCH_TRANSPARENT_HUGEPAGE_PUD.
We should probably push that as a cleanup later and somebody who can
test that config can do that? Currently i can't boot ppc64 with
DBUG_VM_PGTABLE enabled on ppc64 because it is all buggy w.r.t rules.
-aneesh
next prev parent reply other threads:[~2020-08-21 7:15 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-08-19 13:00 [PATCH v2 00/13] mm/debug_vm_pgtable fixes Aneesh Kumar K.V
2020-08-19 13:00 ` [PATCH v2 01/13] powerpc/mm: Add DEBUG_VM WARN for pmd_clear Aneesh Kumar K.V
2020-08-19 13:00 ` [PATCH v2 02/13] powerpc/mm: Move setting pte specific flags to pfn_pte Aneesh Kumar K.V
2020-08-19 13:00 ` [PATCH v2 03/13] mm/debug_vm_pgtable/ppc64: Avoid setting top bits in radom value Aneesh Kumar K.V
2020-08-19 13:00 ` [PATCH v2 04/13] mm/debug_vm_pgtables/hugevmap: Use the arch helper to identify huge vmap support Aneesh Kumar K.V
2020-08-19 16:29 ` kernel test robot
2020-08-20 5:31 ` Aneesh Kumar K.V
2020-08-21 8:38 ` Anshuman Khandual
2020-08-19 13:00 ` [PATCH v2 05/13] mm/debug_vm_pgtable/savedwrite: Enable savedwrite test with CONFIG_NUMA_BALANCING Aneesh Kumar K.V
2020-08-19 13:01 ` [PATCH v2 06/13] mm/debug_vm_pgtable/THP: Mark the pte entry huge before using set_pmd/pud_at Aneesh Kumar K.V
2020-08-19 13:01 ` [PATCH v2 07/13] mm/debug_vm_pgtable/set_pte/pmd/pud: Don't use set_*_at to update an existing pte entry Aneesh Kumar K.V
2020-08-20 14:32 ` Christophe Leroy
2020-08-21 7:14 ` Aneesh Kumar K.V [this message]
2020-08-21 8:20 ` Anshuman Khandual
2020-08-19 13:01 ` [PATCH v2 08/13] mm/debug_vm_pgtable/thp: Use page table depost/withdraw with THP Aneesh Kumar K.V
2020-08-19 13:01 ` [PATCH v2 09/13] mm/debug_vm_pgtable/locks: Move non page table modifying test together Aneesh Kumar K.V
2020-08-19 13:01 ` [PATCH v2 10/13] mm/debug_vm_pgtable/locks: Take correct page table lock Aneesh Kumar K.V
2020-08-21 8:03 ` Anshuman Khandual
2020-08-21 8:08 ` Aneesh Kumar K.V
2020-08-19 13:01 ` [PATCH v2 11/13] mm/debug_vm_pgtable/pmd_clear: Don't use pmd/pud_clear on pte entries Aneesh Kumar K.V
2020-08-19 13:01 ` [PATCH v2 12/13] mm/debug_vm_pgtable/hugetlb: Disable hugetlb test on ppc64 Aneesh Kumar K.V
2020-08-19 13:01 ` [PATCH v2 13/13] mm/debug_vm_pgtable: populate a pte entry before fetching it Aneesh Kumar K.V
2020-08-19 13:45 ` [PATCH v2 00/13] mm/debug_vm_pgtable fixes Aneesh Kumar K.V
2020-08-21 3:33 ` Anshuman Khandual
2020-08-21 4:20 ` Anshuman Khandual
2020-08-21 6:53 ` Aneesh Kumar K.V
2020-08-21 8:01 ` Anshuman Khandual
2020-08-21 8:10 ` Aneesh Kumar K.V
2020-08-21 8:50 ` Aneesh Kumar K.V
2020-08-21 9:09 ` Anshuman Khandual
2020-08-21 8:51 ` Anshuman Khandual
2020-09-01 8:03 ` Christophe Leroy
2020-09-01 9:11 ` Anshuman Khandual
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=3d966519-0a6b-3ccb-fd21-b7f06c8e4df7@linux.ibm.com \
--to=aneesh.kumar@linux.ibm.com \
--cc=akpm@linux-foundation.org \
--cc=anshuman.khandual@arm.com \
--cc=christophe.leroy@csgroup.eu \
--cc=linux-mm@kvack.org \
--cc=linuxppc-dev@lists.ozlabs.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).