diff for duplicates of <20160523214942.GA79646@black.fi.intel.com> diff --git a/a/1.txt b/N1/1.txt index ff0d5fe..ad01121 100644 --- a/a/1.txt +++ b/N1/1.txt @@ -46,7 +46,7 @@ On Mon, May 23, 2016 at 04:13:03PM -0400, Rik van Riel wrote: > I can see a few alternatives here: > > 1) huge pmd collapsing takes both the pmd lock and the pte lock, -> preventing pte updates from happening simultaneously +> preventing pte updates from happening simultaneously That's what we do now and that's not enough. @@ -58,11 +58,11 @@ That's huge hit on scalability. > > 2) code that (re-)acquires the pte lock can read a sequence number -> at the pmd level, check that it did not change after the -> pte lock has been acquired, and abort if it has - I believe most -> of the code that re-acquires the pte lock already knows how to -> abort if somebody else touched the pte while it was looking -> elsewhere +> at the pmd level, check that it did not change after the +> pte lock has been acquired, and abort if it has - I believe most +> of the code that re-acquires the pte lock already knows how to +> abort if somebody else touched the pte while it was looking +> elsewhere So, every pmd_lock() (and other means we take the lock) should bump the sequence number and we need to be able to read stable result outside @@ -79,9 +79,3 @@ And I'm not convinced the hassle worth the gain. -- Kirill A. Shutemov - --- -To unsubscribe, send a message with 'unsubscribe linux-mm' in -the body to majordomo@kvack.org. For more info on Linux MM, -see: http://www.linux-mm.org/ . -Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a> diff --git a/a/content_digest b/N1/content_digest index 0baa4aa..3ab9cf3 100644 --- a/a/content_digest +++ b/N1/content_digest @@ -77,7 +77,7 @@ "> I can see a few alternatives here:\n" "> \n" "> 1) huge pmd collapsing takes both the pmd lock and the pte lock,\n" - "> preventing pte updates from happening simultaneously\n" + "> \302\240 \302\240preventing pte updates from happening simultaneously\n" "\n" "That's what we do now and that's not enough.\n" "\n" @@ -89,11 +89,11 @@ "\n" "> \n" "> 2) code that (re-)acquires the pte lock can read a sequence number\n" - "> at the pmd level, check that it did not change after the\n" - "> pte lock has been acquired, and abort if it has - I believe most\n" - "> of the code that re-acquires the pte lock already knows how to\n" - "> abort if somebody else touched the pte while it was looking\n" - "> elsewhere\n" + "> \302\240 \302\240at the pmd level, check that it did not change after the\n" + "> \302\240 \302\240pte lock has been acquired, and abort if it has - I believe most\n" + "> \302\240 \302\240of the code that re-acquires the pte lock already knows how to\n" + "> \302\240 \302\240abort if somebody else touched the pte while it was looking\n" + "> \302\240 \302\240elsewhere\n" "\n" "So, every pmd_lock() (and other means we take the lock) should bump the\n" "sequence number and we need to be able to read stable result outside\n" @@ -109,12 +109,6 @@ "> considering a race with other pte-level manipulations requires that).\n" "\n" "-- \n" - " Kirill A. Shutemov\n" - "\n" - "--\n" - "To unsubscribe, send a message with 'unsubscribe linux-mm' in\n" - "the body to majordomo@kvack.org. For more info on Linux MM,\n" - "see: http://www.linux-mm.org/ .\n" - "Don't email: <a href=mailto:\"dont@kvack.org\"> email@kvack.org </a>" + Kirill A. Shutemov -11a532e61d1004bc9f287d440d0378a80b6ac7312aee467582579b9cfe0e4303 +38b54409f77097d8e60ff7593777a4b500bdd287816b113cde170dc66c604079
This is an external index of several public inboxes, see mirroring instructions on how to clone and mirror all data and code used by this external index.