Created attachment 1310845 [details]
Problem happens when only stix-fonts is available,
$ java FontTest
But output should be usually:
$ java FontTest
Example steps to reproduce on Fedora:
dd if=/dev/zero of=f26.img bs=1G count=1
sudo mount $PWD/f26.img $PWD/f26.mount
sudo mock -r fedora-26-x86_64 --rootdir=$PWD/f26.mount --init
sudo mock -r fedora-26-x86_64 --rootdir=$PWD/f26.mount --install java-1.8.0-openjdk-devel stix-fonts
sudo cp FontTest.java $PWD/f26.mount
sudo mock -r fedora-26-x86_64 --rootdir=$PWD/f26.mount --shell
rpm -e aajohan-comfortaa-fonts-2.004-7.fc26.noarch ## installed by default
sudo umount $PWD/f26.mount
sudo rmdir $PWD/f26.mount
sudo rm f26.img
I am not 100% sure there is a problem with stix-fonts,
but this comment in sun/font/CompositeFont.java:CompositeFont()
may tell it should check bounds depending on what offset is
/* We will limit the number of slots to 254.
* We store the slot for a glyph id in a byte and we may use one slot
* for an EUDC font, and we may also create a composite
* using this composite as a backup for a physical font.
* So we want to leave space for the two additional slots.
numSlots = (numSlots <= 254) ? numSlots : 254;
Mario, one for you?
(In reply to Andrew John Hughes from comment #2)
> Mario, one for you?
Isn't that a duplicate of https://bugzilla.redhat.com/show_bug.cgi?id=1484079 ?
I made that a dependency of https://bugzilla.redhat.com/show_bug.cgi?id=1484079 since I'm logging work on that bug, as far as I can tell it's the same issue.
(In reply to Mario Torre from comment #3)
> (In reply to Andrew John Hughes from comment #2)
> > Mario, one for you?
> Isn't that a duplicate of
> https://bugzilla.redhat.com/show_bug.cgi?id=1484079 ?
I asked the same in https://bugzilla.redhat.com/show_bug.cgi?id=1484079#c8 ;)
I'm working with upstream for this bug (and 1484079), for now I think we can ship our RPMs with this patch on the next update, since things will likely take longer.
Can you clarify the expectations for when the fix may appear? z-stream? 7.5?
We too are wondering if the date for the anticipated fix has been determined.
We are deploying CSDC's Amanda product on JBOSS 6.4 (RHEL 7.4) and are encountering several locations in the application where functions are failing to execute. CSDC has traced the problem to this Bug so we are eager to get a resolution.
(In reply to Mark Butt from comment #8)
> Hi Mario,
> We too are wondering if the date for the anticipated fix has been determined.
> We are deploying CSDC's Amanda product on JBOSS 6.4 (RHEL 7.4) and are
> encountering several locations in the application where functions are
> failing to execute. CSDC has traced the problem to this Bug so we are eager
> to get a resolution.
The quickest way to resolve would be to install an additional font or replace STIX with a non CFF one, which will make things work already now. As for the patch actually going in the rpm there are few things to still consider but the patch is basically ready and we are working on shipping it, I can't say if this will hit the next update, the best to do is to keep observing this bug report for news.
*** This bug has been marked as a duplicate of bug 1484079 ***