>
> Signed-off-by: Steve Capper
Looks fine as far as I can tell without knowing arm well.
Acked-by: David Gibson
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
_EIP];
> +#elif defined(__arm__)
> + void *pc = (void *)((ucontext_t
> *)uc)->uc_mcontext.fault_address;
> #endif
>
> verbose_printf("SIGSEGV at %p, PC=%p (sig_expected=%p)\n",
--
David Gibson| I'll have my
On Fri, Nov 16, 2012 at 09:38:55AM +0100, telenn barz wrote:
> On Fri, Nov 16, 2012 at 1:13 AM, David Gibson
> wrote:
>
> > On Thu, Nov 15, 2012 at 04:31:44PM -0500, Eric B Munson wrote:
> > > On 2012-11-15 14:44, telenn barz wrote:
> > > > On Thu, Nov 15, 2012
> huge pages still.
In addition I don't think the glibc morecore callback lets us control
the fallback that closely. IIRC, we just fail the morecore, and glibc
falls back on an mmap(). So even on x86 to do partial fallback we'd
need a different (and much more complex) implementa
On Thu, Sep 13, 2012 at 02:27:33AM -0400, Jan Stancek wrote:
>
>
> - Original Message -
> > From: "David Gibson"
> > To: "Jan Stancek"
> > Cc: libhugetlbfs-devel@lists.sourceforge.net, an...@redhat.com
> > Sent: Thursday, 13 Sept
luck. In this case we were unable
to test at all.
It's a bit of a pain, really, because these were realy bugs and they
are still edge cases with interesting complexities and therefore
likely sites for bugs. I'm not sure how to get around this.
--
David Gibson| I
n't
> found a man page for gethugepagesize.)
>
> Have I overlooked documentation for the other 6 functions or are
> these man pages missing? I don't mind writing these man pages if
> they're missing, but I'm asking so I won't duplicate someone else's
&
if (stats.st_uid == uid && (stats.st_mode & S_IRWXU))
> + return true;
> +
> + return false;
> +}
Couldn't you simplify this logic by using access(2)?
--
David Gibson| I'll
though exactly which sizes are
supported varies.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http:
nough
> issue to justify having thp implement a completely new morecore hook.
I really see no reason not to always madvise(), not even as an option.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NO
On Mon, Jun 13, 2011 at 09:06:16AM -0500, Andrew Hastings wrote:
> On 06/08/2011 09:40 PM, Eric B Munson wrote:
> > On Wed, Jun 8, 2011 at 8:50 PM, David Gibson
> > wrote:
[snip]
> Actually, in thinking about this some more I wonder if the
> implementation ought to be using
On Fri, Jun 10, 2011 at 02:54:46PM +0800, bill wrote:
> At 2011-06-10 09:28:52，"David Gibson" wrote:
>
> >On Thu, Jun 09, 2011 at 04:50:50PM +0800, bill wrote:
> >> Hi, guys:)
> >>
> >>
> >> In order to run a program based on huge
m 3.29Mop/s to 9.60Mop/s. That's a significant
> improvement!
Heh, no. 20% is a significant improvement. 200% is a enormous
improvement.
Although the fact that the change is so huge suggests your MMU sucked
rather badly to begin with...
--
David Gibson
7;s most technically advanced content
> authoring tool. Experience the power of Track Changes, Inline Image
> Editing and ensure content is compliant with Accessibility Checking.
> http://p.sf.net/sfu/ephox-dev2dev
> ___
> Libhugetlbfs-devel
indeed, if you're doing thp without madvise, what is the hugetlbfs
morecore actually doing that default morecore wouldn't.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | mini
On Sun, Jun 05, 2011 at 05:01:46PM +0800, Yu Huang wrote:
> 于 2011/6/4 15:22, David Gibson 写道:
> > Uh, you need the -u option to diff. Plain-style diffs are more or
> > less unusable as patches.
> >
> Thanks you! Here is a new patch with -u option:
>
>
>
---
>
> diff -rN libhugetlbfs-2.12-x86/Makefile libhugetlbfs-2.12-mips/Makefile
> 20c20
> < LDSCRIPT_DIST_ELF = elf32ppclinux elf64ppc elf_i386 elf_x86_64
Uh, you need the -u option to diff. Plain-style diffs are more or
less unusable as patches.
--
Davi
ll of the
> comparisons looking for boolean (yes/no) values to ignore case.
>
> Signed-off-by: Eric B Munson
Acked-by: David Gibson
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank
s" or whatever, after this patch it would merely check
that the variable starts with "yes", so "yesterday" would match for
example.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. N
or write
fault on the page we don't clean it but map it without execute
permission. We only perform the icache flush when we get an actual
execute fault on the page.
You will either need to implement similar hacks in ARM, or move the
flushing logic into the generic code.
--
David Gibson
age tables which are shared
between the threads.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
e, you'll need to teach the run_tests script to
automatically run the old-style link tests only on platforms which
support them.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
gt; implementation. If I succeed in doing this, will you accept my
> patch and merge it into the mainline of libhugetlbfs sourcecode?
Sounds like an excellent idea.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, tha
On Wed, Jan 05, 2011 at 06:27:18PM -0600, Andrew Hastings wrote:
> David Gibson wrote:
> > On Tue, Dec 28, 2010 at 05:37:44PM -0600, Andrew Hastings wrote:
> >> libhugetlbfs redefines shmget() so that it can transparently optionally
> >> add the hugepage flag. Unfortuna
of Cray Inc.
I'm mildly surprised there isn't rather more to fix to get
libhugetlbfs to work with static executables. A dynamic executable
was pretty firmly assumed when it was created.
What components of the libary have you tested with a static executable
and this patch?
--
David Gib
the test suite and I
> figured that it wouldn't the last time.
And why would that be a problem? The local scope variable would
override the global and the inline would still work correctly.
--
David Gibson| I'll have my music baroque, and my code
david AT gib
27;s the last, maybe you could change the
library to dynamically allocate space instead of using BSS; then the
libhugetlbfs MORECORE option would move the library's allocations to
hugepage as well as the main executable's.
--
David Gibson| I'll have my music ba
atch makes
the correction.
---
Signed-off-by: David Gibson
diff --git a/tests/gethugepagesizes.c b/tests/gethugepagesizes.c
index 8a6d2d6..860b9a5 100644
--- a/tests/gethugepagesizes.c
+++ b/tests/gethugepagesizes.c
@@ -255,22 +255,29 @@ void cleanup(void)
cleanup_fake_data();
On Thu, Feb 11, 2010 at 02:19:49PM +0100, Jan Engelhardt wrote:
>
> On Thursday 2010-02-11 05:16, David Gibson wrote:
> >On Mon, Feb 08, 2010 at 06:41:24PM +0100, Jan Engelhardt wrote:
> >>
> >> make
> >> %ifarch sparc sparcv9 # that's target
orm it to automake, but would need review whether I've
> matched everything.
Please no. automake is a hideous tangled mess we can really do
without.
--
David Gibson| I'll have my music baroque, and my code
d
Munson
It would probably makes sense to define ALIGN_UP in terms of ALIGN, or
vice versa. Still, doesn't matter that much.
Acked-by: David Gibson
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, t
On Mon, Feb 08, 2010 at 05:14:59PM +, Mel Gorman wrote:
> On Mon, Feb 08, 2010 at 03:24:22PM +1100, David Gibson wrote:
> > On Thu, Feb 04, 2010 at 09:02:18PM +, Eric B. Munson wrote:
> > > elflink.c tests slice boundaries for all architectures, but this is
> > &g
p this check on x86[_64] builds.
Hrm. The x86 versions of hugetlb_slice_end() and so forth are already
supposed to make this check a nop. We shouldn't need to ifdef this
out.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au |
chiefly at the powerpc manifestation, it may or may not also
trip on other archs). Since these testcases will crash current
kernels, we probably don't want to merge this until we can protect
them with a kernel version test, like other dangerous testcases.
Signed-off-by: David Gibson
Index
pers for shm-fork and shm-getraw are removed.
Signed-off-by: David Gibson
Index: libhugetlbfs/tests/run_tests.py
===
--- libhugetlbfs.orig/tests/run_tests.py2009-11-19 15:05:11.352192251
+1100
+++ libhugetlbfs/tests/run_test
a new hugepage port).
Signed-off-by: David Gibson
Index: libhugetlbfs/tests/run_tests.py
===
--- libhugetlbfs.orig/tests/run_tests.py2009-10-02 13:18:48.0
+1000
+++ libhugetlbfs/tests/run_tests.py 2009-10-02 16:
,7 @@
#include "hugetests.h"
/* Override the working version from libhugetlbfs */
-int hugetlbfs_unlinked_fd(void)
+int hugetlbfs_unlinked_fd_for_size(long page_size)
{
return -1;
}
--
David Gibson| I'll have my music baroque, and my code
david AT gibso
#x27;s failing on the system I'm testing right now.
I'll send a patch to fix this if I get to it, but it's not my first
priority right now.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _t
geadm. But I
haven't been around much since hugeadm came into existence. If what
this script does shouldn't be added to hugeadm, it at least looks as
if some of the things it's doing it should do by invoking hugeadm,
rather than directly.
--
David Gibson| I
so at
>least 2 MB in size ]
>
> Thanks in advance for any insight provided ... and happy hacking !
This cannot be done. Huge pages are a hack on the side of the Linux
MM, and can't be used as a substitute for normal page handling.
--
David Gibson| I'l
ectly full and correct build of the testsuite.
Therefore, this patch provides an environment variable to suppress
those warnings, and uses it from the testsuite makefile to get rid of
them in this case.
Signed-off-by: David Gibson
Index: libhugetlbfs/ld.huge
d-off-by: David Gibson
Index: libhugetlbfs/tests/Makefile
===
--- libhugetlbfs.orig/tests/Makefile2009-08-24 14:45:26.0 +1000
+++ libhugetlbfs/tests/Makefile 2009-08-24 15:10:37.0 +1000
@@ -31,7 +31,7 @@ LDFLAGS32
quot;yes")
+ do_test("heap-overflow", HUGETLB_VERBOSE="1", HUGETLB_MORECORE="yes")
# Run the remapping tests' up-front checks
linkhuge_wordsizes = check_linkhuge_tests()
# Original elflink t
shmoverride_linked.c is auto-generated as a symlink during make, but
is not removed during a "make clean". This corrects the error.
Signed-off-by: David Gibson
Index: libhugetlbfs/tests/Makefile
===
--- libhugetlbfs.
arily adjust the relevant
rlimits around these testcases.
Signed-off-by: David Gibson
Index: libhugetlbfs/tests/run_tests.py
===
--- libhugetlbfs.orig/tests/run_tests.py2009-08-20 12:00:15.0
+1000
+++ libhugetlbfs/
od instead of
the manual equivalent.
Signed-off-by: David Gibson
Index: libhugetlbfs/tests/run_tests.py
===
--- libhugetlbfs.orig/tests/run_tests.py2009-08-19 13:18:58.0
+1000
+++ libhugetlbfs/tests/run_tests.py
ually what happens.
Signed-off-by: David Gibson
Index: libhugetlbfs/tests/run_tests.py
===
--- libhugetlbfs.orig/tests/run_tests.py2009-08-19 15:57:59.0
+1000
+++ libhugetlbfs/tests/run_tests.py 2009-0
least.
Getting it to worth with another linker could be anywhere between easy
and impossible depending on whether that linker has suitable options
to alter the segment alignment.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | m
; libhugetlbfs [dmlb2000ts:15538]: INFO: setup_morecore(): heapaddr =
> >>> 0x80e0
> >>> 0
> >>>
> >>>
> >>> So I'm guessing make check is supposed to be run as root? since there
> >>> seems to be mount commands and e
80. Not empty?
> libhugetlbfs [uber-gamer:31049]: INFO: Mapped hugeseg at
> 0x2ac0. Copying 0x220 bytes and 0 extra bytes from
> 0x8008a0...done
> libhugetlbfs [uber-gamer:31047]: INFO: Prepare succeeded
> Failed to map hugepage segment 1: 80-40a0 (errno=12)
>
orrect at run time when
> application is compiled with "pie" flag.
Ah, yes of course. In that case, no, libhugetlbfs does not support
PIE binaries. Patches welcome, I guess...
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.d
faults while mapping
> "writable section" in to hugepage. However without "-pie" flag
> program runs.
I don't think we've ever considered PIEs, so if anything special needs
to be done to support them, then no, they're not supported.
--
David Gibson
ows the user to override this behavior. If -f is specified to
> run_tests.py these potentially dangerous tests will be run.
Thanks for adding this, Eric.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you
On Thu, May 14, 2009 at 08:41:15AM +0100, Eric B. Munson wrote:
> On Wed, 13 May 2009, David Gibson wrote:
>
> > On Wed, May 13, 2009 at 10:35:53AM +0100, Eric B. Munson wrote:
> > > The readahead, madvise, and the fadvise tests can all hang the running
> > > machin
is say something other than just FAIL, as a hint
that the tests aren't actually running, just assuming a result from
the kernel version. "FAIL (assumed)" maybe?
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au |
t came up.
Another one which just occurred to me is to have an option for the
Python, instead of directly running the testcases to generate a really
dumb shell script whose output can then be parsed back into results by
the Python (so you can run the Python test script on a full machine,
shipping just the
y libc functions which would require
accessing the PLT.
This is sufficient on x86 and powerpc. Apparently there's something
different on the MIPS ABI which means that one of the unmapped
addresses is being touched in the black-hole zone. We'll have to work
out exactly what, and why, a
x",
+i, q[i], RANDOM_CONSTANT ^ i);
}
PASS();
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
On Wed, Nov 19, 2008 at 10:19:23AM -0600, Adam Litke wrote:
> On Mon, 2008-11-17 at 11:50 +1100, David Gibson wrote:
> > On Sat, Nov 15, 2008 at 11:56:54AM +1100, David Gibson wrote:
> > > On Fri, Nov 14, 2008 at 10:06:57AM -0600, Adam Litke wrote:
> > > > On Fri, 20
On Wed, Nov 19, 2008 at 10:15:13AM -0600, Adam Litke wrote:
> On Sat, 2008-11-15 at 11:56 +1100, David Gibson wrote:
> > On Fri, Nov 14, 2008 at 10:06:57AM -0600, Adam Litke wrote:
> > > On Fri, 2008-11-14 at 17:51 +1100, David Gibson wrote:
> > > > On Thu, Nov 13,
ut-down embedded distro for
example.
> While a dependency on Python could possibly impact a small number of
> embedded users, I feel the benefits of a more robust set of tests far
> outweigh the downsides of the added dependency.
I suspect you're right, but I really want to have
On Mon, Nov 17, 2008 at 04:50:39PM +, Mel Gorman wrote:
> On Fri, Nov 14, 2008 at 05:51:29PM +1100, David Gibson wrote:
> > On Thu, Nov 13, 2008 at 08:26:12PM +, Adam Litke wrote:
[snip]
> > Hmm. I was about to say I thought changing to Python was a pretty
> > good
On Sat, Nov 15, 2008 at 11:56:54AM +1100, David Gibson wrote:
> On Fri, Nov 14, 2008 at 10:06:57AM -0600, Adam Litke wrote:
> > On Fri, 2008-11-14 at 17:51 +1100, David Gibson wrote:
> > > On Thu, Nov 13, 2008 at 08:26:12PM +, Adam Litke wrote:
> [snip]
> > > The
On Fri, Nov 14, 2008 at 10:06:57AM -0600, Adam Litke wrote:
> On Fri, 2008-11-14 at 17:51 +1100, David Gibson wrote:
> > On Thu, Nov 13, 2008 at 08:26:12PM +, Adam Litke wrote:
[snip]
> > The one big thing that worries me about switching to Python is that it
> > introdu
lf. Making it harder to do a thorough
test of hugepages on J Random embedded board is a non-trivial
drawback.
Not sure which way to go on this one.
--
David Gibson| I'll have my music baroque, and my code
david AT gib
On Wed, Nov 12, 2008 at 12:32:44PM +, Mel Gorman wrote:
> On Tue, Nov 11, 2008 at 10:45:08PM +1100, David Gibson wrote:
> > On Mon, Nov 10, 2008 at 11:55:31AM +, Mel Gorman wrote:
> > > It has been asserted that the name get_huge_pages() implies the API is
> > >
On Wed, Nov 12, 2008 at 12:31:24PM +, Mel Gorman wrote:
> On Tue, Nov 11, 2008 at 10:42:02PM +1100, David Gibson wrote:
> > On Mon, Nov 10, 2008 at 11:55:30AM +, Mel Gorman wrote:
> > > There has been some disagreement on what the exact purpose of
> > > get_hu
would normally be wasted to
> offset the buffer by some random cacheline. Effort is made to select a
> random cache-line to give good average performance.
>
> Signed-off-by: Mel Gorman <[EMAIL PROTECTED]>
Acked-by: David Gibson <[EMAIL PROTECTED]>
--
David Gibson
;[EMAIL PROTECTED]>
> Acked-by: Adam Litke <[EMAIL PROTECTED]>
Apart from the fairly minor naming and description issues discussed in
reply to 0/3:
Acked-by: David Gibson <[EMAIL PROTECTED]>
--
David Gibson| I'll have my music baroque, and my code
david AT
epage malloc()
implementation. It's just that this initial version isn't a very good
implementationm, at least for space-efficiency. We can fix this
deficiency later, if anyone cares enough, without altering the
interface.
--
David Gibson| I'll have my music ba
s() simple and low-level
is why I was dubious about having it fallback to normal pages in the
first place. I may not have expressed this thought very well at the
time, which might account for why our views of the interface had
diverged so greatly this time around.
--
David Gibson
On Fri, Nov 07, 2008 at 10:08:23AM +, Mel Gorman wrote:
> On Fri, Nov 07, 2008 at 09:08:14AM +1100, David Gibson wrote:
> > On Thu, Nov 06, 2008 at 11:19:18AM -0600, Adam Litke wrote:
> > > On Thu, 2008-11-06 at 16:20 +, Mel Gorman wrote:
[snip]
> > > I think it
On Thu, Nov 06, 2008 at 10:49:35AM +, Mel Gorman wrote:
> On Thu, Nov 06, 2008 at 12:06:23PM +1100, David Gibson wrote:
> > On Thu, Nov 06, 2008 at 12:51:54AM +, Mel Gorman wrote:
> > > On Thu, Nov 06, 2008 at 10:15:54AM +1100, David Gibson wrote:
> > > > On W
be introduced and use of
> get_huge_pages deprecated unless you know what you are getting into.
That would be just what I'm suggesting...
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _
ore like anonymous memory (in that the huge
> pages are freed after an unmap and the data will not persist) and that
> another process cannot open the same fd to snoop data.
Uh.. the pages won'tbe freed on unmap, only on close(
On Thu, Nov 06, 2008 at 10:08:27AM -0600, Adam Litke wrote:
> On Thu, 2008-11-06 at 10:15 +1100, David Gibson wrote:
> > On Wed, Nov 05, 2008 at 09:43:40AM -0600, Adam Litke wrote:
> > > On Wed, 2008-11-05 at 14:02 +, Mel Gorman wrote:
> > > > @@ -127,6 +140,34
On Thu, Nov 06, 2008 at 12:51:54AM +, Mel Gorman wrote:
> On Thu, Nov 06, 2008 at 10:15:54AM +1100, David Gibson wrote:
> > On Wed, Nov 05, 2008 at 09:43:40AM -0600, Adam Litke wrote:
> > > On Wed, 2008-11-05 at 14:02 +, Mel Gorman wrote:
> > > > @@ -127,6 +
() is supposed to be the "raw" low-level interface to
just grab some hugepages. Offsetting within the page should be up to
the caller, if it cares.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you.
as to why things aren't working properly, because
you don't see the bogus . directories in ls normally.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_
Removed redundant handling when pagesize argument is zero
>
> Signed-off-by: Adam Litke <[EMAIL PROTECTED]>
Acked-by: David Gibson <[EMAIL PROTECTED]>
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au |
s good to me.
Acked-by: David Gibson <[EMAIL PROTECTED]>
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_
TED]>
Acked-by: David Gibson <[EMAIL PROTECTED]>
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
| _way_ _around_!
http
= hpage_sizes[hpage_sizes_default_idx].pagesize;
> +
> + path = hugetlbfs_find_path_for_size(page_size);
Can't you just pass page_size through here, even if it's zero.
hugetlbfs_find_path_for_size() already knows how to look up the
default page size.
--
David Gibson
impact for tests is that remapping tests that
> > previously received a SIGKILL or SIGBUS when the hugepage pool was too small
> > will instead FAIL with the message "small_bss is not hugepage" or similar.
> >
> > Signed-off-by: Mel Gorman <[EMAIL PROTECT
s are
allocated at. The new way of formatting the tests should be more
robust against changes in pagetable structure, or any newly added
hugepage sizes.
Signed-off-by: David Gibson <[EMAIL PROTECTED]>
---
Paul, please apply for 2.6.28
Index: working-2.6/arch/powerpc/mm/hu
On Thu, Sep 04, 2008 at 04:08:30PM -0500, Jon Tollefson wrote:
> Jon Tollefson wrote:
> > David Gibson wrote:
> >
> >> On Tue, Sep 02, 2008 at 12:12:27PM -0500, Jon Tollefson wrote:
[snip]
> > I have run through the tests twice now with this new patch using a 4k
&
On Wed, Sep 03, 2008 at 05:19:27PM -0500, Jon Tollefson wrote:
> David Gibson wrote:
> > On Tue, Sep 02, 2008 at 12:12:27PM -0500, Jon Tollefson wrote:
> >
> >> David Gibson wrote:
> >>
> >>> When BenH and I were looking at the new code for han
On Tue, Sep 02, 2008 at 08:54:02AM -0500, Adam Litke wrote:
> On Mon, 2008-09-01 at 15:45 +1000, David Gibson wrote:
> > On Fri, Aug 29, 2008 at 09:51:35AM -0500, Adam Litke wrote:
> > > On Fri, 2008-08-29 at 15:42 +1000, David Gibson wrote:
> > > > On Wed, Aug 27,
On Tue, Sep 02, 2008 at 12:12:27PM -0500, Jon Tollefson wrote:
> David Gibson wrote:
> > When BenH and I were looking at the new code for handling 16G pages,
> > we noticed a small bug. It doesn't actually break anything user
> > visible, but it's certainly not th
if (!pmd_none(*pm))
@@ -233,7 +241,7 @@ pte_t *huge_pte_alloc(struct mm_struct *
addr &= hstate->mask;
pg = pgd_offset(mm, addr);
- pu = pud_alloc(mm, pg, addr);
+ pu = hpud_alloc(mm, pg, addr, hstate);
if (pu) {
pm
On Fri, Aug 29, 2008 at 09:19:41AM -0500, Adam Litke wrote:
> On Fri, 2008-08-29 at 12:25 +1000, David Gibson wrote:
> > On Thu, Aug 28, 2008 at 08:51:58AM -0500, Adam Litke wrote:
> > > On Thu, 2008-08-28 at 11:44 +1000, David Gibson wrote:
> > > > On Wed, Aug 27,
On Fri, Aug 29, 2008 at 09:51:35AM -0500, Adam Litke wrote:
> On Fri, 2008-08-29 at 15:42 +1000, David Gibson wrote:
> > On Wed, Aug 27, 2008 at 06:34:41PM +, Adam Litke wrote:
[snip]
> > Hrm. Something about the structure of all this bothers me, but I'm
> > going
On Fri, Aug 29, 2008 at 09:57:10AM -0500, Adam Litke wrote:
> On Fri, 2008-08-29 at 15:43 +1000, David Gibson wrote:
> > On Wed, Aug 27, 2008 at 09:39:29PM +, Adam Litke wrote:
> > > In some cases (such as a distro kernel with backported features)
> > > libhugetlbf
nvironment variable which has a list of those strings to switch the
features.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_
rsion. Possibly a function to change the default pagesize (from
amongst the available ones) at runtime too. Like I say, need to
think about this some more.
--
David Gibson| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au | minimalist, thank you.
On Thu, Aug 28, 2008 at 08:51:58AM -0500, Adam Litke wrote:
> On Thu, 2008-08-28 at 11:44 +1000, David Gibson wrote:
> > On Wed, Aug 27, 2008 at 06:34:54PM +, Adam Litke wrote:
> > > It is now possible for 32 and 64 bit applications to require different
> > > mount
re available. So
rather than attempting to find a single pagesize/mountpoint for each
wordsize, we should instead attempt to run through the whole testsuite
for each pagesize. Then it's just a matter of not running the 32-bit
tests when we're doing a run on pagesize that's too big.
--
What's the state of the support for multiple hugepage sizes / multiple
mountpoints in libhugetlbfs? I remember patches going past, but
AFAICT it's not in the mainline git as yet.
I'd like to do some testing of the kernel multiple size support using
the libhugetlbfs testsuite.
-
On Wed, Aug 20, 2008 at 11:39:54AM +0100, Mel Gorman wrote:
> On (20/08/08 13:59), David Gibson didst pronounce:
> > This patch adds some tracking to the run_tests script which counts the
> > total number of tests, and the number of passes, failures and other
> > results.
This patch adds some tracking to the run_tests script which counts the
total number of tests, and the number of passes, failures and other
results. This is printed as a summary once the script is finished, so
you can see at a glance if the tests ran clean or not.
Signed-off-by: David Gibson
1 - 100 of 425 matches
Mail list logo
