IcedTeaWeb: using java 9 run args ?

Jiri Vanek jvanek at redhat.com
Mon Feb 11 17:18:13 UTC 2019


Hi Laurent!

I had  harm your patch a bit.
I made the location configurable, as I cannot enforce linux distros to have the args file in bin folder.

Still I kept the default of yours - bin.

That naturally went to issue with searching for args file outside the distro world. So I added one
more copy into win/linux static images.

WDYT?


IN addition I had renamed the variables to have common root, and properly used variable instead of
hardcoded filename.

WDYT?

Thanx for patience. Unless you have much more objections, I will push for you tomorrow. (+rust
launchers)

J.
On 2/8/19 4:26 PM, Laurent Bourgès wrote:
> Hi Jiri,
> I easily merge with your latest changes and it builds properly.
> (I noticed build generated the win-deps folder now)
> 
> Here is the updated patch.
> 
> Cheers,
> Laurent
> 
> Le jeu. 7 févr. 2019 à 18:45, Laurent Bourgès <bourges.laurent at gmail.com
> <mailto:bourges.laurent at gmail.com>> a écrit :
> 
>     Jiri,
> 
>     Le jeu. 7 févr. 2019 à 17:12, Jiri Vanek <jvanek at redhat.com <mailto:jvanek at redhat.com>> a écrit :
> 
>         So the cross-build-bats is pushed.
> 
> 
>     Yes I got the notification, thx.
> 
> 
>         I will elaborate on rest tomorrow.
>         I yu adapt your patch to current head, it will be significant help.But is not necessary.
> 
> 
>     I will try merging tomorrow as I stay at home (flu).
> 
>     Thank you too
>     Laurent
> 
> 
>         TYVM for cooperation!
>          J.
> 
> 
>         On 2/7/19 3:47 PM, Laurent Bourgès wrote:
>         > Hi Jiri,
>         > Good to have your feedback.
>         >
>         >
>         > Le jeu. 7 févr. 2019 à 14:27, Jiri Vanek <jvanek at redhat.com <mailto:jvanek at redhat.com>
>         <mailto:jvanek at redhat.com <mailto:jvanek at redhat.com>>> a écrit :
>         >
>         >     Hi Laurent!
>         >
>         >     How are we with this?
>         >
>         >
>         > I prefer you to do its integration as my patch has conflicts with your changes (am) and my
>         changes
>         > are too big so splitting patch is needed, I agree.
>         >
>         >
>         >     Can I commit my "crosscompile" bats patch, and will you adapt?
>         >     Maybe it woudl be better if you first adapt your patch to my crosscompile patch before
>         I push  - to
>         >     test it  abit.
>         >     Or do you suggest any other workflow?
>         >
>         >
>         > Ok for me, I will follow and merge. If anything is wrong or missing, I will propose
>         smaller patches,
>         > sorry.
>         >
>         >
>         >     After crosscompile patch is in, I 'm most likely +1  for pushing your new bat+the work
>         on @arg file
>         >     (wiht some tuning on my side before push).
>         >
>         >
>         > Excellent plan.
>         >
>         >
>         >     Until now I have not found better solution for jigsaw and other params, so also rust
>         launchers will
>         >     do as you do - on jdk9+, will add this @file to java cmd. That will be done as
>         separate changeset
>         >     (by me)
>         >
>         >
>         > Sounds good and easier to do in rust.
>         >
>         > Thanks a lot.
>         >
>         > Laurent
>         >
>         >
>         >     Thanx!
>         >      J.
>         >     On 1/30/19 3:46 PM, Laurent Bourgès wrote:
>         >     > Hi Jiri,
>         >     >
>         >     >     > I finally finished rewriting / unify windows batch script with linux
>         >     >     > shell script.
>         >     >     >
>         >     >     Taht is bat scripting masterpiece. Congratualtions for gaining eternal patience.
>         >     >
>         >     >
>         >     > thanks.
>         >     >  
>         >     >
>         >     >     > Changes:
>         >     >     > - Xnofork is no more needed on windows because argument parsing is working now
>         >     >     > - use run-args + fix classpaths (portable itw with deps)
>         >     >     > - fix CRLF in bat files
>         >     >
>         >     >  
>         >     > One remaining difference: windows batch does not parse deployment.properties (to
>         look up JVM)
>         >
>         >
>         >     This is ok, I do not intend to add it. If anybody wish, anybody can.  And yes,
>         JAVA_HOME can
>         >     help a lot.
>         >     >
>         >     >     The crlg hack is necessary only for args file.  That really have to be another
>         patch for that.
>         >     >     (still we can keep the thread, as many was told here already))
>         >     >
>         >     >
>         >     > I think the contrary, only batch files really expect CRLF line endings. if you edit
>         using
>         >     notepad or
>         >     > if you use echo "text" then the line ending is important.
>         >     > I suppose jvm arg file are well handled by jvm concerning line endings, so it is not
>         required to
>         >     > have CRLF in launcher_run.args.
>         >     >
>         >     > I agree some changes to Makefile.am could be postponed to another patch.
>         >     >  
>         >     >
>         >     >     >
>         >     >     > I tested it on windows 7/10 and it works great.
>         >     >     > I noted windows shortcuts work now, except on JDK12 (maybe related to
>         >     >     > jigsaw classloader ?)
>         >     >
>         >     >     No idea. Adding Joel. to CC. Maybe you will be able to look into it?
>         >     >     Alex quite IcedTea recently, so there is no windows guy around. I will try to
>         get help in
>         >     RH with
>         >     >     windows, otherwise windows "support" of itw will depend purely on community.
>         >     >
>         >     >
>         >     >     >
>         >     >     > I have questions about header variables (too many) and why these need
>         >     >     > to be absolute paths ?
>         >     >     Because in distribution, they ecan be anywhere.
>         >     >
>         >     >     > it could be relative paths if we add ITW_HOME on the top.
>         >     >     > Of course, it depends between DIST or BUNDLE modes ... like for
>         >     >     > dependencies (win/linux_deps folders)
>         >     >     Yes. Bundle mode will be more happy with a bit different setup. Maybe ITW will
>         once grow
>         >     to be more
>         >     >     standalone then distributions frindly.
>         >     >
>         >     >
>         >     > OK, but it means bundle support needs to re-compute path relative to ITW_HOME on
>         linux & windows.
>         >     > In this case, it is not obvious to guess the relative path to a resource from its
>         absolute
>         >     path and
>         >     > lots of path are hard-coded in windows batch.
>         >     > For example:
>         >     > set "BINARY_LOCATION=%ITW_HOME%/bin/%PROGRAM_NAME%.bat"
>         >     >
>         >     > set "SPLASH_LOCATION=%ITW_HOME%/share/icedtea-web/javaws_splash.png" 
>         >     >
>         >     > It is not good, as ITW resource may be moved in Makefile or by installer, so these
>         paths /
>         >     variables
>         >     > must be maintained if any file is renamed or moved (netx.jar -> javaws.jar recently !)
>         >     >
>         >     > Any idea ? It is possible to provide both absolute path & relative paths but it will
>         be too ugly &
>         >     > complicated.
>         >     >  
>         >     >
>         >     >     >
>         >     >     > Please explain how do you plan to make portable (cross-platform) packages ?
>         >     >     Nothing like that!
>         >     >     I will build (and I already am) linux tarablls on linux, and windows tarballs on
>         windows.
>         >     >
>         >     >
>         >     > In my tests, I build on linux with batch generated (wrong paths inside) then create
>         a zip file
>         >     with
>         >     > linux/win deps.
>         >     > It works on windows using bundle mode (ITW_HOME pointing to the script dir).
>         >     > Here is my do-package.sh script (post-build):
>         >     > ITW_HOME=./install/
>         >     > mkdir -p $ITW_HOME/linux-deps-runtime/
>         >     > mkdir -p $ITW_HOME/win-deps-runtime/
>         >     >
>         >     > # add rhino ?
>         >     >
>         >     > # linux:
>         >     > cp tagsoup-1.2.1.jar $ITW_HOME/linux-deps-runtime/
>         >     >
>         >     > # windows:
>         >     > cp tagsoup-1.2.1.jar $ITW_HOME/win-deps-runtime/
>         >     > cp mslinks.jar $ITW_HOME/win-deps-runtime/
>         >     >
>         >     > zip -r itw-1.8-install.zip $ITW_HOME
>         >     >  
>         >     >
>         >     >     I had attached my version of crosscompiler bats on linux. Compared to yours,
>         thay are true on
>         >     >     windows. Not sure how you were handluing this case. Otherwise they loks same
>         asyours. I
>         >     wrote them
>         >     >     in Monday, but then fell down with flue which keeps persisting.
>         >     >
>         >     >
>         >     > Yes it is really close. As I am not an automake expert, I prefer your approach (true
>         on windows).
>         >     > I will merge your changes and adapt my Makefile patch with less changes (LUNCHERS ->
>         LAUNCHERS,
>         >     > run-args, fix CRLF in batch files)
>         >     >  
>         >     >
>         >     >
>         >     >     Also i noticed, that *image* as done on linux, with crosscompile on, is useless on
>         >     windows.  Yes,
>         >     >     you have bats, but windows-deps directory is  corrupoted from linux build.
>         >     >     Whre I wont to keep this state on 1.7, in head, the only possible solution ois
>         to get rid of
>         >     >     linux/windows*deps*dirs and replace them by singledir deps.  WDYT?
>         >     >
>         >     >
>         >     > I can not build on windows, so I am making a cross-platform package on linux (jar
>         files +
>         >     resources
>         >     > + shell launchers).
>         >     > I agree it would be simpler to put dependencies in a single location (jar files are
>         >     cross-platform).
>         >     >  
>         >     >
>         >     >     In addition, yours patch contains also the arg file. I probably agree that it is
>         way to
>         >     go. This
>         >     >     will do its job for shell launchers. For rust ones, I will need to add the
>         cp/bootcp and
>         >     other java
>         >     >     switches customisation properties into deployment.properties. WDYT?
>         >     >
>         >     >
>         >     > The aim consist in having all ITW jigsaw args (add-reads / add-opens) gathered in a
>         single
>         >     place and
>         >     > minimize the jvm args.
>         >     > I wonder if rust launchers could do the same (use @launchers-run.args) and only
>         customize the
>         >     > --patch-moduleargs (netx,plugin,js) and of course, bootclasspath + classpath
>         (absolute paths)
>         >     >  
>         >     >
>         >     >     >
>         >     >     > PS2: you can have a look @github to see complete shell scripts (not diff):
>         >     >     > https://github.com/bourgesl/icedtea-web/tree/run-args/shell-launcher
>         >     >
>         >     >     h<t is really splendid work. I\m unable to have nay comment to new bat file,
>         except that
>         >     it loosk
>         >     >     much better then old one, and that  it do not looks malicious:)
>         >     >
>         >     >
>         >     > Yes I mimic the linux behaviour (functional blocks); only the JVM lookup (from
>         >     deplyoment.properties
>         >     > is not handled).
>         >     >  
>         >     >
>         >     >     I'm for its incorporation to both head and 1.7
>         >     >
>         >     > Great !
>         >     >
>         >     >
>         >     >     Few comments to ther chnegs inline:
>         >     >     ...
>         >     >     >
>         >     >     > patch-run-args-win.2.diff
>         >     >
>         >     >     All refactorings/untypos must go as separate patch. Sorry.  Feel free topush
>         them without
>         >     any other
>         >     >     review. They are making reading of this hard.
>         >     >
>         >     >
>         >     > Please be more precise. You mean fixing typos in Makefile.am (luncher -> launcher) ?
>         >     >  
>         >     >
>         >     >     > +if ENABLE_WIN_SHELL_LAUNCHERS
>         >     >     > +# convert Unix newlines (LF) to DOS format:
>         >     >     > +  line_end_edit_launcher_script=-e "s/\$$/\r/"
>         >     >     > +else
>         >     >     > +  line_end_edit_launcher_script=
>         >     >     > +endif
>         >     >
>         >     >     Only for args file path, right?
>         >     >
>         >     > No, it only fix generated batch files.
>         >     >  
>         >     >
>         >     >     > +if "%MODULAR_JDK%" == "YES" (
>         >     >     > +  rem warning extra escaping
>         >     >     > +  set "MODULAR_ARGS=--patch-module "java.desktop=%NETX_JAR%;%PLUGIN_JAR%""
>         >     >
>         >     >     why NETX.JAR?
>         >     >
>         >     >
>         >     > %NETX_JAR% means $NETX_JAR ie .../javaws.jar
>         >     > It is required to patch jdk.desktop module with ITW code (jigsaw issue as before).
>         >     >
>         >     >  
>         >     >
>         >     >     > +# Support Modular JDK (jigsaw):
>         >     >     >  MODULAR_JDK="NO"
>         >     >     > -version=`${JAVA} -version 2>&1 | head -n 1 | cut -d'-' -f1 | cut -d'"' -f2 |
>         cut -d'.' -f1`
>         >     >     > +fullversion=`${JAVA} -version 2>&1`
>         >     >     > +echo "fullversion: $fullversion"
>         >     >
>         >     >     I would skipp the fullversion: string, and printed $fullversion to stderr as are
>         people
>         >     used from
>         >     >     jdk/ (?)
>         >     >
>         >     >
>         >     > Not understood. I wanted to avoid calling java -version twice and but show JVM
>         version once.
>         >     > What do you propose ?
>         >     >  
>         >     >
>         >     >     > 
>         >     >     > +#echo "CMD: ${COMMAND[@]}"
>         >     >
>         >     >     The commented echo is weird. One is usually running this in -x subshell to see
>         details.
>         >     >
>         >     >
>         >     > Ok I will remove such debugging lines.
>         >     >
>         >     > PS: Would it sound possible for you to handle merging & pushing this patch ?
>         >     > Anyway I will wait for your changes to Makefiles before merging.
>         >     >
>         >     > Cheers,
>         >     > Laurent
>         >
>         >
>         >     --
>         >     Jiri Vanek
>         >     Senior QE engineer, OpenJDK QE lead, Mgr.
>         >     Red Hat Czech
>         >     jvanek at redhat.com <mailto:jvanek at redhat.com> <mailto:jvanek at redhat.com
>         <mailto:jvanek at redhat.com>>    M: +420775390109
>         >
> 
> 
>         -- 
>         Jiri Vanek
>         Senior QE engineer, OpenJDK QE lead, Mgr.
>         Red Hat Czech
>         jvanek at redhat.com <mailto:jvanek at redhat.com>    M: +420775390109
> 
> 
> 
> -- 
> -- 
> Laurent Bourgès


-- 
Jiri Vanek
Senior QE engineer, OpenJDK QE lead, Mgr.
Red Hat Czech
jvanek at redhat.com    M: +420775390109
-------------- next part --------------
A non-text attachment was scrubbed...
Name: patch-run-args-win.4.diff
Type: text/x-patch
Size: 33285 bytes
Desc: not available
URL: <https://mail.openjdk.java.net/pipermail/distro-pkg-dev/attachments/20190211/be920f13/patch-run-args-win.4-0001.diff>


More information about the distro-pkg-dev mailing list