NativeLibLoader - installLibraryFromResource - RuntimeImage
Scott Palmer
swpalmer at gmail.com
Fri Jul 27 15:07:00 UTC 2018
> On Jul 27, 2018, at 10:49 AM, Kevin Rushforth <kevin.rushforth at oracle.com> wrote:
>
> Those are all excellent points. This suggests we might want to adopt a general policy of "the current OpenJFX dev works with the latest released OpenJDK plus the current OpenJDK under development" is the better way to go. Otherwise it will be several years before OpenJFX can adopt, say, a new JDK 12 or JDK 13 feature.
>
> I note that this matches the model and spirit of the OpenJDK development. If you want to get the latest and greatest features, use the latest feature release. If you want stability, use a supported LTS release. Makes sense to me.
Which can be much easier said than done. I generally like to live near the bleeding edge… I *want to switch*, but we are still stuck on Java 8 (like I suspect most companies are) because of the very significant development effort to move to Java 9 or beyond. Java 9 module's access controls were one thing, Java 11’s removal of some modules and development tools further broke the world. Not to mention that good module support in the tool chain is still not available. We are waiting for the dust to settle, we need to wait for many dependencies and tools to catch up and become Java 11 compatible, then we have to try to migrate our own code and hope that all the dependency upgrades didn’t break things. Basically we won’t even be able to start to migrate our code beyond Java 8 until next year (when Java 8 LTS ends).
I’ve already spent days trying to get code to build with Java 11 (unsuccessfully). It’s the first time in the history of Java that I haven’t been able to upgrade the JRE with trivial effort. Usually the only thing holding me back was waiting for Proguard to support the new class version. (And I could at least develop and test without obfuscation until it was ready.)
I am hoping that once we make the leap from Java 8 to Java 11, that further updates will be less painful. If so, then I could agree that the latest OpenJFX can use the latest released JRE.
Unless you want to maintain a LTS release of OpenJFX in addition to the “current” version?
Scott
>
> -- Kevin
>
>
> On 7/27/2018 7:35 AM, Johan Vos wrote:
>> I think we have to distinguish between the technical work done by OpenJFX and the distributions/support offered by companies; similar to the technical work in OpenJDK and the distributions/support offered by Oracle and others.
>>
>> The technical work done here should imo be able to use all functionality offered by the latest released JDK (11/12/...)
>>
>> I don't think there is a good reason on why the head-development of OpenJFX can not use the latest JDK release. If developers/companies want to switch to the latest and greatest OpenJFX release, they can also upgrade to the latest released OpenJDK.
>>
>> And if there are company rules preventing this, well, that is a reason to consider commercial support.
>> I don't like to use my commercial hat in this list, but this is why we have JavaFX Enterprise support with Gluon: https://gluonhq.com/commercial-support-for-javafx/
>> Also, not that it is relevant on this list, but the revenues from commercial enterprise support allow us to spend resources on working on OpenJFX.
>>
>> - Johan
>>
>> On Fri, Jul 27, 2018 at 4:12 PM Kevin Rushforth <kevin.rushforth at oracle.com <mailto:kevin.rushforth at oracle.com><mailto:kevin.rushforth at oracle.com <mailto:kevin.rushforth at oracle.com>>> wrote:
>>
>> Hi Steve,
>>
>>
>>> I don't know if this discussion already comes up but I didn't
>>> found any related answers.
>>> Is there a self-defined goal about the compability of versions
>>> OpenJFX and the OracleJDK LTS versions?
>>> E.g. OpenJFX will be compatible to OracleJDK 11 until the next
>>> LTS version of Oracle will be released
>>
>> That's an excellent question. While I might expect OpenJFX 12 to
>> stop running on JDK 10 at some point (JDK 10 is EOLed by JDK 11,
>> and dropping support for it would allow us to get rid of the old
>> swing interop implementation and build logic), having it run on
>> JDK 11 is a requirement: At a minimum we will need to run on the
>> latest released version at any point, since we are no longer
>> bundled. Beyond that, it's up for discussion among the OpenJFX
>> community: I can see value in allowing, say, OpenJFX 13 or 14 to
>> continue to run with JDK 11 LTS, but there is also a cost to doing
>> so, both in terms of testing and also in either avoiding new JDK
>> features or using reflection or other techniques to optionally use
>> them.
>>
>> What do others think?
>>
>>
>> -- Kevin
>>
>>
>>
>> On 7/27/2018 6:43 AM, Steve Hruda wrote:
>>> Hi Kevin,
>>> at the moment it's not really a use-case. I didn't used the jmods
>>> because I test if the maven artifacts could easier our life in
>>> case of maintaining different product versions.
>>> I love the easy distribution of the maven artifacts across
>>> different machines (developers & build servers).
>>>
>>> But so far I'm not sure if I the maven artifacts bring me a real
>>> benefit compared to the jmod's.
>>>
>>> Off-Topic:
>>> I don't know if this discussion already comes up but I didn't
>>> found any related answers.
>>> Is there a self-defined goal about the compability of versions
>>> OpenJFX and the OracleJDK LTS versions?
>>> E.g. OpenJFX will be compatible to OracleJDK 11 until the next
>>> LTS version of Oracle will be released
>>>
>>> Best Regards,
>>> Steve
>>> Kevin Rushforth <kevin.rushforth at oracle.com <mailto:kevin.rushforth at oracle.com>
>>> <mailto:kevin.rushforth at oracle.com <mailto:kevin.rushforth at oracle.com>>> schrieb am Do., 26. Juli
>>> 2018, 14:42:
>>>
>>> Hi Steve,
>>>
>>> Seems it will be an easy bug for Johan to fix, but I'm
>>> curious about how
>>> you created your custom image. The jmods are more suitable
>>> for that
>>> approach, since their purpose is to be used with jlink to
>>> create a
>>> custom image, in which case the native libraries will be
>>> linked into the
>>> image. Maybe I'm missing some use case, though.
>>>
>>> -- Kevin
>>>
>>>
>>> On 7/26/2018 2:46 AM, Johan Vos wrote:
>>> > Hi Steve,
>>> >
>>> > That looks like a bug. The libPrefix and libSuffix are
>>> indeed not set in
>>> > cases where usingModules is true, which is only the case
>>> when the jrt
>>> > protocol is used.
>>> > It seems to me the prefix/suffix should always be computed.
>>> It doesn't look
>>> > right that they are computed inside the
>>> loadLibraryFullPath. But it looks
>>> > even worse that the "reallib" is calculated based on
>>> statements that might
>>> > not be reached, where it could have used
>>> System.mapLibraryName()
>>> >
>>> > I'll create an issue and a PR to fix this.
>>> >
>>> > Thanks for reporting,
>>> >
>>> > - Johan
>>> >
>>> >
>>> > On Thu, Jul 26, 2018 at 11:03 AM Steve Hruda
>>> <steve.hruda at gmail.com <mailto:steve.hruda at gmail.com> <mailto:steve.hruda at gmail.com <mailto:steve.hruda at gmail.com>>> wrote:
>>> >
>>> >> Hi,
>>> >> I created a custom runtime image (windows x64) for my
>>> JavaFX application
>>> >> and used the OpenJFX 11-ea+19 maven artifacts.
>>> >>
>>> >> I get the following exception if I try to execute my
>>> application:
>>> >>
>>> >> Graphics Device initialization failed for : d3d, sw
>>> >> Error initializing QuantumRenderer: no suitable pipeline found
>>> >> java.lang.RuntimeException: java.lang.RuntimeException:
>>> Error initializing
>>> >> QuantumRenderer: no suitable pipeline found
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumRenderer.getInstance(QuantumRenderer.java:280)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumToolkit.init(QuantumToolkit.java:222)
>>> >> at
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.Toolkit.getToolkit(Toolkit.java:260)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:263)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:157)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.startToolkit(LauncherImpl.java:658)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplicationWithArgs(LauncherImpl.java:409)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplication(LauncherImpl.java:363)
>>> >> at
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
>>> >> Method)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
>>> >> at
>>> java.base/java.lang.reflect.Method.invoke(Method.java:564)
>>> >> at
>>> >>
>>> >>
>>> java.base/sun.launcher.LauncherHelper$FXHelper.main(LauncherHelper.java:941)
>>> >> Caused by: java.lang.RuntimeException: Error initializing
>>> QuantumRenderer:
>>> >> no suitable pipeline found
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumRenderer$PipelineRunnable.init(QuantumRenderer.java:94)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumRenderer$PipelineRunnable.run(QuantumRenderer.java:124)
>>> >> at java.base/java.lang.Thread.run(Thread.java:844)
>>> >> Exception in thread "main"
>>> java.lang.reflect.InvocationTargetException
>>> >> at
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
>>> >> Method)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
>>> >> at
>>> java.base/java.lang.reflect.Method.invoke(Method.java:564)
>>> >> at
>>> >>
>>> >>
>>> java.base/sun.launcher.LauncherHelper$FXHelper.main(LauncherHelper.java:941)
>>> >> Caused by: java.lang.RuntimeException: No toolkit found
>>> >> at
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.Toolkit.getToolkit(Toolkit.java:272)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:263)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:157)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.startToolkit(LauncherImpl.java:658)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplicationWithArgs(LauncherImpl.java:409)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplication(LauncherImpl.java:363)
>>> >> ... 5 more
>>> >>
>>> >>
>>> >> -Djavafx.verbose=true shows that JavaFX isn't able to load
>>> the native libs
>>> >> from the javafx-graphics jar.
>>> >>
>>> >>
>>> >> I debugged the NativeLibLoader and loadLibraryFromResource and
>>> >> installLibraryFromResource will be executed which is ok.
>>> But the
>>> >> *reallib *value for
>>> >> all dll's is wrong because of the empty *libSuffix*.
>>> >>
>>> >> e.g. the reallib value of *api-ms-win-core-console-l1-1-0* is
>>> >> */api-ms-win-core-console-l1-1-0* and not
>>> >> */api-ms-win-core-console-l1-1-0.dll*
>>> >>
>>> >> Is it intentional that *libPrefix *& *libSuffix *will not
>>> be set in case of
>>> >> the jrt protocol ?
>>> >>
>>> >>
>>> >>
>>> https://github.com/javafxports/openjdk-jfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L308 <https://github.com/javafxports/openjdk-jfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L308>
>>> >>
>>> >>
>>> >>
>>> >>
>>> >>
>>> https://github.com/javafxports/openjdkjfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L338 <https://github.com/javafxports/openjdkjfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L338>
>>> >>
>>> >>
>>> >>
>>> >>
>>> >> Best Regards,
>>> >> Steve
>>> >>
>>>
>>>
>>> Kevin Rushforth <kevin.rushforth at oracle.com <mailto:kevin.rushforth at oracle.com>
>>> <mailto:kevin.rushforth at oracle.com <mailto:kevin.rushforth at oracle.com>>> schrieb am Do., 26. Juli
>>> 2018, 14:42:
>>>
>>> Hi Steve,
>>>
>>> Seems it will be an easy bug for Johan to fix, but I'm
>>> curious about how
>>> you created your custom image. The jmods are more suitable
>>> for that
>>> approach, since their purpose is to be used with jlink to
>>> create a
>>> custom image, in which case the native libraries will be
>>> linked into the
>>> image. Maybe I'm missing some use case, though.
>>>
>>> -- Kevin
>>>
>>>
>>> On 7/26/2018 2:46 AM, Johan Vos wrote:
>>> > Hi Steve,
>>> >
>>> > That looks like a bug. The libPrefix and libSuffix are
>>> indeed not set in
>>> > cases where usingModules is true, which is only the case
>>> when the jrt
>>> > protocol is used.
>>> > It seems to me the prefix/suffix should always be computed.
>>> It doesn't look
>>> > right that they are computed inside the
>>> loadLibraryFullPath. But it looks
>>> > even worse that the "reallib" is calculated based on
>>> statements that might
>>> > not be reached, where it could have used
>>> System.mapLibraryName()
>>> >
>>> > I'll create an issue and a PR to fix this.
>>> >
>>> > Thanks for reporting,
>>> >
>>> > - Johan
>>> >
>>> >
>>> > On Thu, Jul 26, 2018 at 11:03 AM Steve Hruda
>>> <steve.hruda at gmail.com <mailto:steve.hruda at gmail.com> <mailto:steve.hruda at gmail.com <mailto:steve.hruda at gmail.com>>> wrote:
>>> >
>>> >> Hi,
>>> >> I created a custom runtime image (windows x64) for my
>>> JavaFX application
>>> >> and used the OpenJFX 11-ea+19 maven artifacts.
>>> >>
>>> >> I get the following exception if I try to execute my
>>> application:
>>> >>
>>> >> Graphics Device initialization failed for : d3d, sw
>>> >> Error initializing QuantumRenderer: no suitable pipeline found
>>> >> java.lang.RuntimeException: java.lang.RuntimeException:
>>> Error initializing
>>> >> QuantumRenderer: no suitable pipeline found
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumRenderer.getInstance(QuantumRenderer.java:280)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumToolkit.init(QuantumToolkit.java:222)
>>> >> at
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.Toolkit.getToolkit(Toolkit.java:260)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:263)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:157)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.startToolkit(LauncherImpl.java:658)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplicationWithArgs(LauncherImpl.java:409)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplication(LauncherImpl.java:363)
>>> >> at
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
>>> >> Method)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
>>> >> at
>>> java.base/java.lang.reflect.Method.invoke(Method.java:564)
>>> >> at
>>> >>
>>> >>
>>> java.base/sun.launcher.LauncherHelper$FXHelper.main(LauncherHelper.java:941)
>>> >> Caused by: java.lang.RuntimeException: Error initializing
>>> QuantumRenderer:
>>> >> no suitable pipeline found
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumRenderer$PipelineRunnable.init(QuantumRenderer.java:94)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.quantum.QuantumRenderer$PipelineRunnable.run(QuantumRenderer.java:124)
>>> >> at java.base/java.lang.Thread.run(Thread.java:844)
>>> >> Exception in thread "main"
>>> java.lang.reflect.InvocationTargetException
>>> >> at
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native
>>> >> Method)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
>>> >> at
>>> >>
>>> >>
>>> java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
>>> >> at
>>> java.base/java.lang.reflect.Method.invoke(Method.java:564)
>>> >> at
>>> >>
>>> >>
>>> java.base/sun.launcher.LauncherHelper$FXHelper.main(LauncherHelper.java:941)
>>> >> Caused by: java.lang.RuntimeException: No toolkit found
>>> >> at
>>> >>
>>> javafx.graphics/com.sun.javafx.tk.Toolkit.getToolkit(Toolkit.java:272)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:263)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.PlatformImpl.startup(PlatformImpl.java:157)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.startToolkit(LauncherImpl.java:658)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplicationWithArgs(LauncherImpl.java:409)
>>> >> at
>>> >>
>>> >>
>>> javafx.graphics/com.sun.javafx.application.LauncherImpl.launchApplication(LauncherImpl.java:363)
>>> >> ... 5 more
>>> >>
>>> >>
>>> >> -Djavafx.verbose=true shows that JavaFX isn't able to load
>>> the native libs
>>> >> from the javafx-graphics jar.
>>> >>
>>> >>
>>> >> I debugged the NativeLibLoader and loadLibraryFromResource and
>>> >> installLibraryFromResource will be executed which is ok.
>>> But the
>>> >> *reallib *value for
>>> >> all dll's is wrong because of the empty *libSuffix*.
>>> >>
>>> >> e.g. the reallib value of *api-ms-win-core-console-l1-1-0* is
>>> >> */api-ms-win-core-console-l1-1-0* and not
>>> >> */api-ms-win-core-console-l1-1-0.dll*
>>> >>
>>> >> Is it intentional that *libPrefix *& *libSuffix *will not
>>> be set in case of
>>> >> the jrt protocol ?
>>> >>
>>> >>
>>> >>
>>> https://github.com/javafxports/openjdk-jfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L308 <https://github.com/javafxports/openjdk-jfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L308>
>>> >>
>>> >>
>>> >>
>>> >>
>>> >>
>>> https://github.com/javafxports/openjdkjfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L338 <https://github.com/javafxports/openjdkjfx/blob/c168ab56accd7e74d53737bc0832495dbc318e52/modules/javafx.graphics/src/main/java/com/sun/glass/utils/NativeLibLoader.java#L338>
>>> >>
>>> >>
>>> >>
>>> >>
>>> >> Best Regards,
>>> >> Steve
>>> >>
More information about the openjfx-dev
mailing list