Thread (38 messages) 38 messages, 9 authors, 2018-10-16

Re: [PATCH v2 2/2] mm: speed up mremap by 500x on large regions

From: David Miller <davem@davemloft.net>
Date: 2018-10-12 18:02:27
Also in: kvmarm, linux-alpha, linux-mips, linux-mm, linux-riscv, linux-sh, linux-um, lkml

From: "Kirill A. Shutemov" <redacted>
Date: Fri, 12 Oct 2018 14:30:56 +0300
I looked into the code more and noticed move_pte() helper called from
move_ptes(). It changes PTE entry to suite new address.

It is only defined in non-trivial way on Sparc. I don't know much about
Sparc and it's hard for me to say if the optimization will break anything
there.

I think it worth to disable the optimization if __HAVE_ARCH_MOVE_PTE is
defined. Or make architectures state explicitely that the optimization is
safe.
What sparc is doing in move_pte() is flushing the data-cache
(synchronously) if the virtual address color of the mapping changes.

Hope this helps.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help