We want to document the possible need to use --relax with GNU ld to avoid
truncated relocations. Update the documentation.
Signed-off-by: Adam Litke <[EMAIL PROTECTED]>
---
HOWTO | 10 ++
1 files changed, 10 insertions(+), 0 deletions(-)
diff --git a/HOWTO b/HOWTO
index 54519e2..7de9
On Tue, Oct 28, 2008 at 09:44:10AM -0500, Adam Litke wrote:
> On Tue, 2008-10-28 at 13:54 +, Andy Whitcroft wrote:
> > When testing the library we sometimes want to influence data which becomes
> > cached within the library. To allow this we need to reset that cached data
> > to ensure it is r
On Tue, 28 Oct 2008, Andy Whitcroft wrote:
> When testing the library we sometimes want to influence data which becomes
> cached within the library. To allow this we need to reset that cached data
> to ensure it is re-read from our test data instead of the real system.
> This change adds the conc
On Tue, 2008-10-28 at 13:54 +, Andy Whitcroft wrote:
> When testing some gethugepagesizes we desire to fake the default huge
> page size, to test both having and not having hugepage support on the
> same machine. Now that the default huge page size is cached we now
> need to invalidate this ca
On Tue, 2008-10-28 at 13:54 +, Andy Whitcroft wrote:
> When testing the library we sometimes want to influence data which becomes
> cached within the library. To allow this we need to reset that cached data
> to ensure it is re-read from our test data instead of the real system.
> This change
On Tue, 2008-10-28 at 13:54 +, Andy Whitcroft wrote:
> We use the kernel default page size all over the place. We use it
> approximatly 6 times when setting the boundaries on a single pool.
> Currently every place looks this value up in /proc/meminfo which is
> very costly.
>
> As this value
最廉价！最高效的产品推广方式！
A，我公司专业代帮你把你的网站或商品上传到国内外如:百度雅虎,Google,阿里巴
巴,,,等大中小各类网站的不同栏目上，信息有效期5年以上。
（注：每成功上传1万个网站仅收费50元）
B，我公司专业代发电子邮件,可把你的商品信息轻松发进2,4亿企业或个人的信箱！
(注：每成功代发10万邮件仅收费20元）。
C，我公司拥有2,4亿中外邮箱地址出售(分国家,省区,行业).
（注：每10万地址售价20元）
联系信息：
陕西省汉中市甲天下软件公司
陕西省汉中市兴汉路168号附一号
电话：0916-225
On Tue, 28 Oct 2008, Andy Whitcroft wrote:
> When testing some gethugepagesizes we desire to fake the default huge
> page size, to test both having and not having hugepage support on the
> same machine. Now that the default huge page size is cached we now
> need to invalidate this cache to allow
The first patch in this series has been seen before, it optimises for the
common "lookup the default huge page size" operation. However this then
breaks the gethugepagesizes test as it can no longer manipulate the values
read as they are cached in the library.
The second patch implements a new te
When testing some gethugepagesizes we desire to fake the default huge
page size, to test both having and not having hugepage support on the
same machine. Now that the default huge page size is cached we now
need to invalidate this cache to allow us test with various values.
Add and expose kernel_
We use the kernel default page size all over the place. We use it
approximatly 6 times when setting the boundaries on a single pool.
Currently every place looks this value up in /proc/meminfo which is
very costly.
As this value is essentially static per boot we can cache this value.
Create a new
When testing the library we sometimes want to influence data which becomes
cached within the library. To allow this we need to reset that cached data
to ensure it is re-read from our test data instead of the real system.
This change adds the concept of a test probe point to the main library.
These
This patch Adjusts the HOWTO section on sizing huge page pools to reference
the hugeadm utlity.
Also it changes the number of pages from 20 to 25 for most test cases.
Signed-off-by: Eric B Munson <[EMAIL PROTECTED]>
---
HOWTO | 28 +---
1 files changed, 25 insertions(+)
NNew liffe!
http://cid-d5b4140c3621e66a.spaces.live.com/blog/cns!D5B4140C3621E66A!114.entry
Do you think his arm will ever be all right? She to see the
close. Hasten to save the queenand in a large way. Mr. Pallzey
feared socialism and king ordered a saddle brought, and
was placing the
14 matches
Mail list logo
