You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@commons.apache.org by "Dapeng Sun (JIRA)" <ji...@apache.org> on 2016/04/28 05:33:12 UTC

[jira] [Updated] (CRYPTO-31) Make fields final wherever possible

     [ https://issues.apache.org/jira/browse/CRYPTO-31?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Dapeng Sun updated CRYPTO-31:
-----------------------------
    Affects Version/s: 1.0.0

> Make fields final wherever possible
> -----------------------------------
>
>                 Key: CRYPTO-31
>                 URL: https://issues.apache.org/jira/browse/CRYPTO-31
>             Project: Commons Crypto
>          Issue Type: Improvement
>          Components: Stream
>    Affects Versions: 1.0.0
>            Reporter: Sebb
>            Assignee: Ferdinand Xu
>            Priority: Minor
>             Fix For: 1.0.0
>
>
> Final fields are automatically thread safe if the object itself is not modified.
> Whereas non-final fields are not necessarily thread-safe even if they are not mutated - this is because of the Java memory model.
> There are several non-final fields that are not mutated once created. They should be made final to improve thread-safety. This is particularly important for static fields since they are more likely to be used by multiple threads.
> Note: such changes will not affect the public API if the fields are private or package-protected.
> The fields include:
> JavaSecureRandom.instance
> OpensslSecureRandom.fallback
> OpensslSecureRandom.nativeEnabled
> ChannelInput.channel
> StreamInput.buf - since no point creating an instance if it is not going to be used for reading, might as well create the buffer up front
> StreamOutput.bufferSize
> StreamOutput.buf - as for StreamInput
> StreamOutput.out - should also be private
> static NativeCodeLoader.nativeCodeLoaded
> static OSInfo.archMapping
> static ReflectionUtils.classLoader



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

RE: [jira] [Updated] (CRYPTO-31) Make fields final wherever possible

Posted by "Sun, Dapeng" <da...@intel.com>.
Hi Gary,

Thank you for pointing it out. I will remove the tag of 'affects version'.

Regards
Dapeng

-----Original Message-----
From: Gary Gregory [mailto:garydgregory@gmail.com] 
Sent: Thursday, April 28, 2016 1:03 PM
To: Commons Developers List
Subject: Re: [jira] [Updated] (CRYPTO-31) Make fields final wherever possible

I do not think tickets should be marked as affects 1.0.0 since that version is not released.

Gary
On Apr 27, 2016 8:33 PM, "Dapeng Sun (JIRA)" <ji...@apache.org> wrote:

>
>      [
> https://issues.apache.org/jira/browse/CRYPTO-31?page=com.atlassian.jir
> a.plugin.system.issuetabpanels:all-tabpanel
> ]
>
> Dapeng Sun updated CRYPTO-31:
> -----------------------------
>     Affects Version/s: 1.0.0
>
> > Make fields final wherever possible
> > -----------------------------------
> >
> >                 Key: CRYPTO-31
> >                 URL: https://issues.apache.org/jira/browse/CRYPTO-31
> >             Project: Commons Crypto
> >          Issue Type: Improvement
> >          Components: Stream
> >    Affects Versions: 1.0.0
> >            Reporter: Sebb
> >            Assignee: Ferdinand Xu
> >            Priority: Minor
> >             Fix For: 1.0.0
> >
> >
> > Final fields are automatically thread safe if the object itself is 
> > not
> modified.
> > Whereas non-final fields are not necessarily thread-safe even if 
> > they
> are not mutated - this is because of the Java memory model.
> > There are several non-final fields that are not mutated once created.
> They should be made final to improve thread-safety. This is 
> particularly important for static fields since they are more likely to 
> be used by multiple threads.
> > Note: such changes will not affect the public API if the fields are
> private or package-protected.
> > The fields include:
> > JavaSecureRandom.instance
> > OpensslSecureRandom.fallback
> > OpensslSecureRandom.nativeEnabled
> > ChannelInput.channel
> > StreamInput.buf - since no point creating an instance if it is not 
> > going
> to be used for reading, might as well create the buffer up front
> > StreamOutput.bufferSize
> > StreamOutput.buf - as for StreamInput StreamOutput.out - should also 
> > be private static NativeCodeLoader.nativeCodeLoaded static 
> > OSInfo.archMapping static ReflectionUtils.classLoader
>
>
>
> --
> This message was sent by Atlassian JIRA
> (v6.3.4#6332)
>

Re: [jira] [Updated] (CRYPTO-31) Make fields final wherever possible

Posted by Gary Gregory <ga...@gmail.com>.
I do not think tickets should be marked as affects 1.0.0 since that version
is not released.

Gary
On Apr 27, 2016 8:33 PM, "Dapeng Sun (JIRA)" <ji...@apache.org> wrote:

>
>      [
> https://issues.apache.org/jira/browse/CRYPTO-31?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
> ]
>
> Dapeng Sun updated CRYPTO-31:
> -----------------------------
>     Affects Version/s: 1.0.0
>
> > Make fields final wherever possible
> > -----------------------------------
> >
> >                 Key: CRYPTO-31
> >                 URL: https://issues.apache.org/jira/browse/CRYPTO-31
> >             Project: Commons Crypto
> >          Issue Type: Improvement
> >          Components: Stream
> >    Affects Versions: 1.0.0
> >            Reporter: Sebb
> >            Assignee: Ferdinand Xu
> >            Priority: Minor
> >             Fix For: 1.0.0
> >
> >
> > Final fields are automatically thread safe if the object itself is not
> modified.
> > Whereas non-final fields are not necessarily thread-safe even if they
> are not mutated - this is because of the Java memory model.
> > There are several non-final fields that are not mutated once created.
> They should be made final to improve thread-safety. This is particularly
> important for static fields since they are more likely to be used by
> multiple threads.
> > Note: such changes will not affect the public API if the fields are
> private or package-protected.
> > The fields include:
> > JavaSecureRandom.instance
> > OpensslSecureRandom.fallback
> > OpensslSecureRandom.nativeEnabled
> > ChannelInput.channel
> > StreamInput.buf - since no point creating an instance if it is not going
> to be used for reading, might as well create the buffer up front
> > StreamOutput.bufferSize
> > StreamOutput.buf - as for StreamInput
> > StreamOutput.out - should also be private
> > static NativeCodeLoader.nativeCodeLoaded
> > static OSInfo.archMapping
> > static ReflectionUtils.classLoader
>
>
>
> --
> This message was sent by Atlassian JIRA
> (v6.3.4#6332)
>