<AWT Dev>  Review request for 8074824: Resolve disabled warnings for libawt_xawt
semyon.sadetsky at oracle.com
Tue Oct 17 16:34:40 UTC 2017
Thank you for the information. I ran SwingSet under GTK2. No issues found.
On 10/16/2017 09:41 PM, Ajit Ghaisas wrote:
> Hi Semyon,
> I had attempted this fix some time back and had run into some regression for SwingSet2.
> There was a crash while using SwingSet2 with GTK look-and-feel with GTK-2.
> I do not remember the exact steps, but just playing with SwingSet2 had produced this crash.
> Can you please check your fix does not cause such a regression?
> -----Original Message-----
> From: Sergey Bylokhov
> Sent: Tuesday, October 17, 2017 5:15 AM
> To: Semyon Sadetsky; Phil Race; awt-dev at openjdk.java.net
> Subject: Re: <AWT Dev>  Review request for 8074824: Resolve disabled warnings for libawt_xawt
> Hi, Semyon.
> A few notes:> I ran OGL tests and found a typo. Below the update webrev> http://cr.openjdk.java.net/~ssadetsky/8074824/webrev.01/
> - What is a typo? It seems the v0 and v1 are identical.
> - The new "IS_SAFE_SIZE_T" macro returns different result than the old one for "0"
> - What is the reason to inline some of the shaders in OGLPaints.c and leave as-is others? For example "noCycleCode" was inlined and "texCoordCalcCode" still stored in the "static const char *". It seems that the "static const" style is more readable.
> Best regards, Sergey.
More information about the awt-dev