Hi Guys,

I've provided a gcc-specific fix in the makefile to prevent the warning.

-- Awt2dLibraries.gmk:471 --
DISABLED_WARNINGS_gcc := array-bounds clobbered implicit-fallthrough 
shift-negative-value, \

I've also provided an underflow fix in the .c file to fix the problem 
*causing* the warning.

-- jchuff.c:808 --
while ((bits[j] == 0) && (j > 0))

Either will work fine.

Note: After determining that it affects multiple gcc versions, and that 
the logic to make a makefile do a 
compare (the shell business) on the gcc version seemed hacky to me, I 
considered the best solution
to be one of the two simple fixes outlined above. This seemed to be 
acceptable to people in the
community, yet we're still having trouble getting this fix through. 

I'm not sure why.

Best Regards 

Adam Farley


> Hi Phil!
> 
> 
> thanks for pointing out the history, I was not aware of that.
> 
> 
> I looked at that huffman coding and tried to determine whether the 
underflow may happen in real life scenarios. I could at least not exclude 
that possibility. I looked thru the mailing list threads - did someone 
analyse and conclude for sure this was just a pointless compiler warning?
> 
> 
> I would prefer the pragmatic solution (and IMHO also safer one) of 
fixing this underflow in the proposed fashion. I had opened a bug report 
earlier today. However, if someone already spent brain cycles on it and a 
patch - in whatever form - is forthcoming, I do not want to butt in. In 
that case I will close this bug again. 
> 
> 
> I would just like to see this fixed this because it affects us at SAP 
too.
> 
> 
> Kind Regards, Thomas
> 
> 
> On Wed, Mar 21, 2018 at 6:56 PM, Phil Race <philip.r...@oracle.com> 
wrote:
> 
> I prefer the makefile fix, since we don't by policy, make changes to the 
imported libraries.
> 
> On Jan 23rd [1] I expressed such a tool-chain specific makefile fix 
would be fine by me.
> 
> Toolchain specific means ideally it would look like what Magnus wrote 
[2]
> 
> Although you said GC 5.4.0 would need to be included in the logic.
> 
> If it can be shown to affect current / future versions of gcc then it 
could be unqualified.
> 
> I think we've just been waiting for a webrev since then ..
> 
> -phil.
> 
> [1] 
http://mail.openjdk.java.net/pipermail/2d-dev/2018-January/008855.html
> [2] 
http://mail.openjdk.java.net/pipermail/build-dev/2018-January/020695.html
> 
> 
> 
> On 03/21/2018 09:53 AM, Adam Farley8 wrote:
> 
> :) 
> 
> > Hi Adam, 
> > 
> > no problem. I'll open a bug and if necessary find a second reviewer. 
Thanks for fixing, maybe I can stop building with warnings disabled on our 
s390 machines now. 
> > 
> > ..Thomas 
> > 
> > > On Wed, Mar 21, 2018 at 5:10 PM, Andrew Leonard 
<andrew_m_leon...@uk.ibm.com> wrote: 
> > > Hi Thomas, 
> > > I'm a "contributor", but not a "committer", so not on that list, 
didn't even know that 
> > > list existed! I was sort of assuming since it was a trivial change, 
and the request was 
> > > for a review, i'd chip in...! 
> > > Thanks 
> > > Andrew 
> > 
> > > Andrew Leonard 
> > > Java Runtimes Development 
> > > IBM Hursley 
> > > IBM United Kingdom Ltd 
> > > Phone internal: 245913, external: 01962 815913 
> > > internet email: andrew_m_leon...@uk.ibm.com 
> > 
> 

Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 
741598. 
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU

Reply via email to