Andrea Arcangeli <[EMAIL PROTECTED]> wrote:
>
> you're right something might be different now
> that we don't follow a swapout virtual address space order anymore
There's a patch in -mm which attempts to do so, and afair, succeeds.
However the performance seems to be crappy. Its main benefit
On Mon, Feb 14, 2005 at 06:36:43PM +, Hugh Dickins wrote:
> On Mon, 14 Feb 2005, Andrea Arcangeli wrote:
> > > By the way, while we're talking of remove_exclusive_swap_page:
> > > a more functional issue I sometimes wonder about, why don't we
> > > remove_exclusive_swap_page on write fault?
On Mon, 14 Feb 2005, Andrea Arcangeli wrote:
> > By the way, while we're talking of remove_exclusive_swap_page:
> > a more functional issue I sometimes wonder about, why don't we
> > remove_exclusive_swap_page on write fault? Keeping the swap slot
> > is valuable if read fault, but once the page
On Mon, 14 Feb 2005, Andrea Arcangeli wrote:
By the way, while we're talking of remove_exclusive_swap_page:
a more functional issue I sometimes wonder about, why don't we
remove_exclusive_swap_page on write fault? Keeping the swap slot
is valuable if read fault, but once the page is
On Mon, Feb 14, 2005 at 06:36:43PM +, Hugh Dickins wrote:
On Mon, 14 Feb 2005, Andrea Arcangeli wrote:
By the way, while we're talking of remove_exclusive_swap_page:
a more functional issue I sometimes wonder about, why don't we
remove_exclusive_swap_page on write fault? Keeping the
Andrea Arcangeli [EMAIL PROTECTED] wrote:
you're right something might be different now
that we don't follow a swapout virtual address space order anymore
There's a patch in -mm which attempts to do so, and afair, succeeds.
However the performance seems to be crappy. Its main benefit at
On Fri, 11 Feb 2005, Andrea Arcangeli wrote:
>
> Ok, I'm quite convinced it's correct now. The only thing that can make
> mapcount go up without the lock on the page without userspace
> intervention (and userspace intervention would make it an undefined
> behaviour like in my example with fork),
On Fri, Feb 11, 2005 at 07:23:09AM +, Hugh Dickins wrote:
> And it's fine for the behaviour to be somewhat undefined in this
> peculiar case: the important thing is just that the page must not
> be freed and reused while I/O occurs, hence get_user_page raising
> the page_count - which I'm
On Fri, Feb 11, 2005 at 07:23:09AM +, Hugh Dickins wrote:
And it's fine for the behaviour to be somewhat undefined in this
peculiar case: the important thing is just that the page must not
be freed and reused while I/O occurs, hence get_user_page raising
the page_count - which I'm _not_
On Fri, 11 Feb 2005, Andrea Arcangeli wrote:
Ok, I'm quite convinced it's correct now. The only thing that can make
mapcount go up without the lock on the page without userspace
intervention (and userspace intervention would make it an undefined
behaviour like in my example with fork), was
On Thu, 10 Feb 2005, Andrea Arcangeli wrote:
> On Thu, Feb 10, 2005 at 08:19:47PM +, Hugh Dickins wrote:
> >
> > get_user_pages broke COW in advance of the I/O. The reason for
> > subsequent COW if the page gets unmapped is precisely because
> > can_share_swap_page used page_count to judge
On Thu, Feb 10, 2005 at 08:19:47PM +, Hugh Dickins wrote:
> On Thu, 10 Feb 2005, Andrea Arcangeli wrote:
> >
> > The reason pinned pages cannot be unmapped is that if they're being
> > unmapped while a rawio read is in progress, a cow on some shared
> > swapcache under I/O could happen.
> >
On Thu, 10 Feb 2005, Andrea Arcangeli wrote:
>
> The reason pinned pages cannot be unmapped is that if they're being
> unmapped while a rawio read is in progress, a cow on some shared
> swapcache under I/O could happen.
>
> If a try_to_unmap + COW over a shared swapcache happens during a rawio
>
On Thu, Feb 10, 2005 at 11:16:44AM -0800, Dave Hansen wrote:
> We actually do that, in addition to the more active methods of capturing
> the memory that we're about to remove.
This sounds the best way to go. btw, is this a different codebase from
the one that Toshihiro is testing?
> You're
On Thu, 2005-02-10 at 20:05 +0100, Andrea Arcangeli wrote:
> Overall I see nothing wrong in preventing memory removal while rawio is
> in flight. rawio cannot be in flight forever (ptrace and core dump as
> well should complete eventually). Why can't you simply drop pages from
> the freelist one
On Tue, Feb 08, 2005 at 04:26:26PM +, Hugh Dickins wrote:
> Seems it was okay after all, I got confused by an unrelated issue.
> Here's what I had in mind, what do you think? Does it indeed help
> with your problem? I'm copying Andrea because it was he who devised
> that fix to the
At Tue, 8 Feb 2005 16:26:26 + (GMT),
Hugh Dickins wrote:
>
> On Mon, 7 Feb 2005, Hugh Dickins wrote:
> > On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
> > > The current implementation of memory hotremoval relies on that pages
> > > can be unmapped from process spaces. After successful
At Tue, 8 Feb 2005 16:26:26 + (GMT),
Hugh Dickins wrote:
On Mon, 7 Feb 2005, Hugh Dickins wrote:
On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
The current implementation of memory hotremoval relies on that pages
can be unmapped from process spaces. After successful unmapping,
On Tue, Feb 08, 2005 at 04:26:26PM +, Hugh Dickins wrote:
Seems it was okay after all, I got confused by an unrelated issue.
Here's what I had in mind, what do you think? Does it indeed help
with your problem? I'm copying Andrea because it was he who devised
that fix to the
On Thu, 2005-02-10 at 20:05 +0100, Andrea Arcangeli wrote:
Overall I see nothing wrong in preventing memory removal while rawio is
in flight. rawio cannot be in flight forever (ptrace and core dump as
well should complete eventually). Why can't you simply drop pages from
the freelist one by
On Thu, Feb 10, 2005 at 11:16:44AM -0800, Dave Hansen wrote:
We actually do that, in addition to the more active methods of capturing
the memory that we're about to remove.
This sounds the best way to go. btw, is this a different codebase from
the one that Toshihiro is testing?
You're right,
On Thu, 10 Feb 2005, Andrea Arcangeli wrote:
The reason pinned pages cannot be unmapped is that if they're being
unmapped while a rawio read is in progress, a cow on some shared
swapcache under I/O could happen.
If a try_to_unmap + COW over a shared swapcache happens during a rawio
read,
On Thu, Feb 10, 2005 at 08:19:47PM +, Hugh Dickins wrote:
On Thu, 10 Feb 2005, Andrea Arcangeli wrote:
The reason pinned pages cannot be unmapped is that if they're being
unmapped while a rawio read is in progress, a cow on some shared
swapcache under I/O could happen.
If a
On Thu, 10 Feb 2005, Andrea Arcangeli wrote:
On Thu, Feb 10, 2005 at 08:19:47PM +, Hugh Dickins wrote:
get_user_pages broke COW in advance of the I/O. The reason for
subsequent COW if the page gets unmapped is precisely because
can_share_swap_page used page_count to judge
At Mon, 7 Feb 2005 21:24:59 + (GMT),
Hugh Dickins wrote:
>
> On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
> > The current implementation of memory hotremoval relies on that pages
> > can be unmapped from process spaces. After successful unmapping,
> > subsequent accesses to the pages are
At Mon, 7 Feb 2005 21:24:59 + (GMT),
Hugh Dickins wrote:
On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
The current implementation of memory hotremoval relies on that pages
can be unmapped from process spaces. After successful unmapping,
subsequent accesses to the pages are blocked and
On Mon, 7 Feb 2005, Hugh Dickins wrote:
> On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
> > The current implementation of memory hotremoval relies on that pages
> > can be unmapped from process spaces. After successful unmapping,
> > subsequent accesses to the pages are blocked and don't interfere
On Mon, 7 Feb 2005, Hugh Dickins wrote:
On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
The current implementation of memory hotremoval relies on that pages
can be unmapped from process spaces. After successful unmapping,
subsequent accesses to the pages are blocked and don't interfere
the
On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
> The current implementation of memory hotremoval relies on that pages
> can be unmapped from process spaces. After successful unmapping,
> subsequent accesses to the pages are blocked and don't interfere
> the hotremoval operation.
>
> However, this
On Thu, 3 Feb 2005, IWAMOTO Toshihiro wrote:
The current implementation of memory hotremoval relies on that pages
can be unmapped from process spaces. After successful unmapping,
subsequent accesses to the pages are blocked and don't interfere
the hotremoval operation.
However, this code
The current implementation of memory hotremoval relies on that pages
can be unmapped from process spaces. After successful unmapping,
subsequent accesses to the pages are blocked and don't interfere
the hotremoval operation.
However, this code
if (PageSwapCache(page) &&
The current implementation of memory hotremoval relies on that pages
can be unmapped from process spaces. After successful unmapping,
subsequent accesses to the pages are blocked and don't interfere
the hotremoval operation.
However, this code
if (PageSwapCache(page)
32 matches
Mail list logo