Hi Holmes,
I agree with your comment on duplicated comment. It should be
factored out.
The context of the change is that the group of signal handlers should
be registered at best effort, which means for all JVM, we should try to
register them all in an independent way. In fact, the origin
On 6/08/2012 6:57 PM, Frank Ding wrote:
Hi Holmes,
I agree with your comment on duplicated comment. It should be factored out.
The context of the change is that the group of signal handlers should be
registered at best effort, which means for all JVM, we should try to
register them all in an inde
- Original Message -
> Andrew,
>
> I just pushed 7188852. Is it possible for you to update your webrev
> with
> this change? I would
> like to try a full build (and regression tests) on your final bits to
> make sure we don't break
> the build.
>
Do you mean the ZLIB_LIBS changes? I alr
- Original Message -
> Andrew,
>
> I just pushed 7188852. Is it possible for you to update your webrev
> with
> this change? I would
> like to try a full build (and regression tests) on your final bits to
> make sure we don't break
> the build.
>
Do you mean the ZLIB_LIBS changes? I alr
- Original Message -
>
> On 8/5/2012 2:00 PM, Alan Bateman wrote:
> > On 03/08/2012 19:33, Andrew Hughes wrote:
> >> :
> >> http://cr.openjdk.java.net/~andrew/syslibs/zlib/webrev.02/
> >>
> >> is an updated version which checks if ZLIB_LIBS is set on Solaris,
> >> GNU/Linux
> >> and MacO
Hi Andrew,
I meant if we are going to put SYSTEM_ZLIB=true as default for linux as
Alan suggested,
we might need to update the build document as well to include zlib-dev
as the "necessary"
package to build jdk on linux.
Currently the SYSTEM_ZLIB=true is set in make/common/Defs-macosx.gmk.
ZLI
- Original Message -
> Hi Andrew,
>
> I meant if we are going to put SYSTEM_ZLIB=true as default for linux
> as
> Alan suggested,
> we might need to update the build document as well to include
> zlib-dev
> as the "necessary"
> package to build jdk on linux.
Ok. I think that's a separate
On 08/06/2012 11:26 AM, Andrew Hughes wrote:
- Original Message -
Hi Andrew,
I meant if we are going to put SYSTEM_ZLIB=true as default for linux
as
Alan suggested,
we might need to update the build document as well to include
zlib-dev
as the "necessary"
package to build jdk on linux.
Thanks Andrew. 7189533 created for this.
David
-
On 3/08/2012 8:03 AM, Andrew Hughes wrote:
- Original Message -
Hi Andrew,
On 2/08/2012 7:35 PM, Andrew Hughes wrote:
- Original Message -
Andrew et al,
AFAICS here:
220 encoding_variant = malloc(strlen(te
Hi Holmes,
It is not only when -Xrs is specified. The circumstances can be
illustrated below
1. a JVM has its own extra signal registration that occupies say, "INT",
prior to the execution Terminator.setup().
2. User customizes HotSpot to have similar overridden signal handling.
In both ca
On 7/08/2012 1:03 PM, Frank Ding wrote:
Hi Holmes,
David :)
It is not only when -Xrs is specified. The circumstances can be
illustrated below
1. a JVM has its own extra signal registration that occupies say, "INT",
prior to the execution Terminator.setup().
2. User customizes HotSpot to have
11 matches
Mail list logo
