patchset 2 report (a love story)

patchset 2 report (a love story)

Post by Brent Vern » Thu, 06 Feb 2003 21:10:52



Hi,

  After a long 6 hour compile that ended in a failure on a plugin
target, I had a fresh native 1.4.1 image (with HotSpot no less!)
for my 4.7-stable box.  I've been running a couple of tomcat apps
pretty heavily for the past few hours, and things seem stable, aside
from a few DEBUG: signal messages[1].  Aside from apparent stability,
the performance improvement over native 1.3.1 is about 600%.  
Nice work.

  Now, a nit... I tried building on a -current machine, but the
3.2.1 compiler bails early in the build of adlc, complaining of
numerous ostream operator link-errors.  I hacked around with
this a bit, while the 4.7 box was building, but gave up after
I noticed that I could compile the same ostream<<(char) code
that the build was complaining about.  Weird.

  In any case, kudos Alexey, for the fine work.  If there is any
legwork that I could do in getting 3.2.1 compiling on -current,
let me know; I'd be glad to help.

  I created a ports directory for the jdk14, and was wondering if
it would be possible to distribute patchset 2 in .zip format, so
USE_ZIP could be set in the Makefile to extract both the scsl-src
and patchset, instead of having to _manually_ extract one of them.
If I'm missing some feature of the ports system that would allow
automatic extraction of both, please throw that cluestick at me :-)

  On the HotSpot note, is there any work being done on getting 1.3.1's
HotSpot working?  AFAIR, it would not build with 1.3.1-patchset-7.

cheers.
  brent

[1] one of these is signal 11, so if this is _the_ signal 11?I'm
    familiar with, and you want to know how to duplicate, I could
    probably distill a testcase.

--
"Develop your talent, man, and leave the world something. Records are
really gifts from people. To think that an artist would love you enough
to share his music with anyone is a beautiful thing."  -- Duane Allman


with "unsubscribe freebsd-java" in the body of the message

 
 
 

patchset 2 report (a love story)

Post by Alexey Zelk » Thu, 06 Feb 2003 22:56:07


hi,


> Hi,

>   After a long 6 hour compile that ended in a failure on a plugin

Not good, plugin build is completely disabled for now...  I'll take a
look.

BTW, if you've fast machine with enough memory you may try to build
hotspot object files parallely:

setenv HOTSPOT_BUILD_JOBS 4

before starting the build (for example)

Quote:> target, I had a fresh native 1.4.1 image (with HotSpot no less!)

Sorry, there's no support for something other than HotSpot in JDK 1.4 :-)

Quote:> for my 4.7-stable box.  I've been running a couple of tomcat apps
> pretty heavily for the past few hours, and things seem stable, aside
> from a few DEBUG: signal messages[1].  Aside from apparent stability,
> the performance improvement over native 1.3.1 is about 600%.  
> Nice work.

Cool!

Quote:>   Now, a nit... I tried building on a -current machine, but the
> 3.2.1 compiler bails early in the build of adlc, complaining of
> numerous ostream operator link-errors.  I hacked around with
> this a bit, while the 4.7 box was building, but gave up after
> I noticed that I could compile the same ostream<<(char) code
> that the build was complaining about.  Weird.

Strange.  Can you please provide me information about exact error
messages, 'gcc -v', and 'uname -a' ?

Quote:>   In any case, kudos Alexey, for the fine work.  If there is any
> legwork that I could do in getting 3.2.1 compiling on -current,
> let me know; I'd be glad to help.

Provide me details on failure, please.

Quote:>   I created a ports directory for the jdk14, and was wondering if
> it would be possible to distribute patchset 2 in .zip format, so
> USE_ZIP could be set in the Makefile to extract both the scsl-src
> and patchset, instead of having to _manually_ extract one of them.
> If I'm missing some feature of the ports system that would allow
> automatic extraction of both, please throw that cluestick at me :-)

I'd like to look.  I think Greg too :)

Quote:>   On the HotSpot note, is there any work being done on getting 1.3.1's
> HotSpot working?  AFAIR, it would not build with 1.3.1-patchset-7.

Yes.  Patchset 8 for 1.3.1 will appear soon...  Greg ?

Quote:> [1] one of these is signal 11, so if this is _the_ signal 11?I'm
>     familiar with, and you want to know how to duplicate, I could
>     probably distill a testcase.

It's debug printf, signal 11 is indicating that something goes wierd
inside java compiled code.  But since you don't see any exception or
fatal failures -- JVM is able to recover from this problem.


with "unsubscribe freebsd-java" in the body of the message

 
 
 

patchset 2 report (a love story)

Post by Alexey Zelk » Fri, 07 Feb 2003 03:06:28


hi,

[useful link & descriptions]

Thanks!

Quote:>   Makefile and distinfo assume presence of a zipped version of the
> patchset, and the *-for-tgz pieces use the tarball patchset.  I'll
> try to build with WITHOUT_PLUGIN to see if I can get the pkg-plist
> in a state where it can be used to build a package.

Try to set 'NO_PLUGIN' to 'YES' before build and let me know result.


with "unsubscribe freebsd-java" in the body of the message

 
 
 

patchset 2 report (a love story)

Post by Brent Vern » Fri, 07 Feb 2003 03:02:08


[2003-02-05 16:01] Alexey Zelkin said:

| >   Now, a nit... I tried building on a -current machine, but the
| > 3.2.1 compiler bails early in the build of adlc, complaining of
| > numerous ostream operator link-errors.  I hacked around with
| > this a bit, while the 4.7 box was building, but gave up after
| > I noticed that I could compile the same ostream<<(char) code
| > that the build was complaining about.  Weird.
|
| Strange.  Can you please provide me information about exact error
| messages, 'gcc -v', and 'uname -a' ?




Using built-in specs.
Configured with: FreeBSD/i386 system compiler
Thread model: posix
gcc version 3.2.1 [FreeBSD] 20021119 (release)

Output from the failed make is at:

  http://rcfile.org/~brent/jdk14/make.out.gz

| >   I created a ports directory for the jdk14, and was wondering if
| > it would be possible to distribute patchset 2 in .zip format, so
| > USE_ZIP could be set in the Makefile to extract both the scsl-src
| > and patchset, instead of having to _manually_ extract one of them.
| > If I'm missing some feature of the ports system that would allow
| > automatic extraction of both, please throw that cluestick at me :-)
|
| I'd like to look.  I think Greg too :)

  While I've not dealt with updating pkg-plist (in fact, I'll have
to go RTFM to create/modify that file) the ports directory
that I built from on the 4.7-stable machine is at:

  http://rcfile.org/~brent/jdk14/jdk14-bsd/

  Makefile and distinfo assume presence of a zipped version of the
patchset, and the *-for-tgz pieces use the tarball patchset.  I'll
try to build with WITHOUT_PLUGIN to see if I can get the pkg-plist
in a state where it can be used to build a package.

cheers.
  brent

--
"Develop your talent, man, and leave the world something. Records are
really gifts from people. To think that an artist would love you enough
to share his music with anyone is a beautiful thing."  -- Duane Allman


with "unsubscribe freebsd-java" in the body of the message

 
 
 

patchset 2 report (a love story)

Post by Bill Huey (H » Fri, 07 Feb 2003 07:19:35



>   After a long 6 hour compile that ended in a failure on a plugin
> target, I had a fresh native 1.4.1 image (with HotSpot no less!)
> for my 4.7-stable box.  I've been running a couple of tomcat apps
> pretty heavily for the past few hours, and things seem stable, aside
> from a few DEBUG: signal messages[1].  Aside from apparent stability,
> the performance improvement over native 1.3.1 is about 600%.  
> Nice work.

That's good to hear. Why do you think the performance is much better in
this system verses 1.3.1 ?

Quote:>   On the HotSpot note, is there any work being done on getting 1.3.1's
> HotSpot working?  AFAIR, it would not build with 1.3.1-patchset-7.

I've been doing other work, but I'm probably going to start up work
again on this matter. I'd like to hand this work over to somebody in
some formal way so that I can move onto other projects and still have
things be completely functional. There is a tiny framework I'm doing is
going to allow for both HotSpot versions, 1.3.x and 1.4.x to run properly
without the use of signal driven ucontext manipulations for both -stable
and -current.

That'll be good for all of the FreeBSD folks and Alexey can then revert
the removal of some critical asserts to the GC system concerning
safepoint timeout.

Dan Eischen's will have to commit the kernel/UTS related ucontext stuff
(don't know what's happened to that) before we can use the signal driven
stuff in -current without modification. Currently that's all turned off
since it's broken. That'll be the proper interface to a threads ucontext
until the thread control structure is pubically available in some formal
capacity when KSEs get into place.

Quote:> [1] one of these is signal 11, so if this is _the_ signal 11?I'm
>     familiar with, and you want to know how to duplicate, I could
>     probably distill a testcase.

Pester me so that I'll do this work soon. I've been in RTOS land for a
while and having really felt pressure to do any more HotSpot work from
anything external to me. :)

bill


with "unsubscribe freebsd-java" in the body of the message

 
 
 

patchset 2 report (a love story)

Post by Greg Lew » Fri, 07 Feb 2003 17:51:13



>   I created a ports directory for the jdk14, and was wondering if
> it would be possible to distribute patchset 2 in .zip format, so
> USE_ZIP could be set in the Makefile to extract both the scsl-src
> and patchset, instead of having to _manually_ extract one of them.
> If I'm missing some feature of the ports system that would allow
> automatic extraction of both, please throw that cluestick at me :-)

I'll look into this.  It sounds like things are pretty positive so
it may be that now is the time to create a port for jdk14 :).

--

Eyes Beyond                         Web     : http://www.eyesbeyond.com


with "unsubscribe freebsd-java" in the body of the message

 
 
 

patchset 2 report (a love story)

Post by Greg Lew » Fri, 07 Feb 2003 17:54:11



> >   On the HotSpot note, is there any work being done on getting 1.3.1's
> > HotSpot working?  AFAIR, it would not build with 1.3.1-patchset-7.

> Yes.  Patchset 8 for 1.3.1 will appear soon...  Greg ?

Yes.  This was derailed for a couple days due to a power supply failing,
but hopefully will get back on track and be out in the next few days.

--

Eyes Beyond                         Web     : http://www.eyesbeyond.com


with "unsubscribe freebsd-java" in the body of the message

 
 
 

patchset 2 report (a love story)

Post by Brent Vern » Sat, 08 Feb 2003 02:02:48


[2003-02-06 19:14] Greg Lewis said:

| >   I created a ports directory for the jdk14, and was wondering if
| > it would be possible to distribute patchset 2 in .zip format, so
| > USE_ZIP could be set in the Makefile to extract both the scsl-src
| > and patchset, instead of having to _manually_ extract one of them.
| > If I'm missing some feature of the ports system that would allow
| > automatic extraction of both, please throw that cluestick at me :-)
|
| I'll look into this.  It sounds like things are pretty positive so
| it may be that now is the time to create a port for jdk14 :).

FWIW, my latest round of changes to my experimental port Makefile
resulted in a successful installation on my 4.7-stable system :-)

  http://www.rcfile.org/~brent/jdk14/jdk14-bsd/

notable changes from the jdk1.3.1p7 port, on which this was based:
  1) updated MAKE_ENV
  2) defined WITHOUT_PLUGIN in Makefile
  3) removed references to non-existent 'images' target
  5) install to jdk-${VERION}-bsd
  4) changed name of port to 'jdk-${VERSION}-bsd-${PATCH_REV}'

problems:
  1) pkg-plist is still not updated.

cheers.
  b

--
"Develop your talent, man, and leave the world something. Records are
really gifts from people. To think that an artist would love you enough
to share his music with anyone is a beautiful thing."  -- Duane Allman


with "unsubscribe freebsd-java" in the body of the message

 
 
 

1. Two hosts and a disk array - a love story :)

Hi!

I need to connect two hosts (Enterprise 2 and a Enterprise 4000) to a
single disk array (model 100).

The disk array has 6 2Gb disks. I have used the Veritas Disk Manager
to create about 6 volumes in several configurations: 3 RAID5, 2
Striped, 1 Mirroring + striping.

I want to be able to access some volumes from the E2 and some from the
E4000.

I know that the DiskArray can be connected to both host. What I could
not find out is if I can use the disk array simultaneous from both
host. Note that no volume will be accessed simultaneouslly from both
host. Each volume will be mounted in a single host.

Thanks for any input on this

Melo
--

* TELENET, Servicos de Telecomunicacoes, SA   -  Tel. +351 1 3871010 *

************* http://www.co.telenet.pt/personal/melo/ ****************

2. mail program for X

3. patchset 2 report (billh's resignation)

4. XFree86 on Apollo9

5. [Fwd: Re: patchset 2 report (billh's resignation)]

6. 8139too on Proliant

7. OT: Patchset website (was: Re: 1.3.1 patchset 8)

8. Setting up Net Gateway in Linux?

9. Microsoft Annual Reports Tell True Linux Story

10. For Unix programmers, who love people and who loves serving them ...

11. Aw or DOOL-Soap love =Real life love

12. syslogd reporting ip address when reporting remotely

13. Database Reporting Tool e.g. Crystal Report