I finally had a chance to finish looking at this problem. It's a buffer
overflow in pkgdata. The pkgdata program is fast and loose with C
strings. It copies stuff all over the place into buffers without
checking lengths. Coincidentally, someone else just also found this and
reported it upstrea
Cyril Brulebois wrote:
> Jay Berkenbilt (17/05/2010):
>> The logs seem to show that it is failing in the same place while
>> running pkgdata. Where are you seeing that it breaks on a different
>> file every time?
>
> As I said, when I reproduced it on asdfasdf.
Oh, okay. I see that too now --
Hi.
Jay Berkenbilt (17/05/2010):
> The logs seem to show that it is failing in the same place while
> running pkgdata. Where are you seeing that it breaks on a different
> file every time?
As I said, when I reproduced it on asdfasdf.
> In any case, pkgdata is part of ICU itself, so I should b
Sorry I overlooked this bug report. Somehow I missed the email and just
found it now in my periodic check on qa.debian.org.
Cyril Brulebois wrote:
> your package FTBFS reliably on kfreebsd-amd64 in experimental, with bus
> errors while trying to build stuff under source/data. I've also
> repro
Source: icu
Version: 4.4.1-1
Severity: serious
Justification: FTBFS
User: debian-...@lists.debian.org
Hi,
your package FTBFS reliably on kfreebsd-amd64 in experimental, with bus
errors while trying to build stuff under source/data. I've also
reproduced those failures in a kfreebsd-amd64 sid chroo
5 matches
Mail list logo
