[PATCH] Robust basic_filebuf::imbue

Paolo Carlini paolo.carlini@oracle.com
Tue Oct 12 11:08:00 GMT 2010


On 10/12/2010 12:21 PM, David Krauss wrote:
> Yikes, there was some very unfortunate commenting in those testcases!
>
> Furthermore, when I uncommented that, one of them failed! But I don't think the bug is in the current code, it's preexisting in basic_filebuf::seekoff, and looks related to something I was already aware of and wanting to fix. It is not returning the correct state with the position in unbuffered mode. So, for now I'm "adding a regression." Will add a fix in a separate patchfile after sleep.
If this is a smaller issue, self-contained in seekoff, can we fix it first?

Paolo.



More information about the Libstdc++ mailing list