On 1/21/07, seth vidal <skvidal@xxxxxxxxxxxxxx> wrote:
Thanks for the patch. The only thing I worry about is the more-or-less global exception catch in your try, except clauses. Is it possible to narrow the set of possible exceptions down there so we don't end up catching the world?
Fair enough. I don't understand how exceptions should be propagating, so here's another take on this bug. Whenever an exception reaches UpdateInstallThread.run, we first release the locks held logically by this thread, and then let the exception propagate as before. The attached patch does not change what exceptions are produced or consumed, but only releases locks before a thread holding these locks is about to terminate. I've tested the patch, and even though UpdateInstallThread keeps halting when the IndexError exception gets thrown, the yum lock is properly released. Nickolai.
Attachment:
yumupd.patch2
Description: Binary data
_______________________________________________ Yum mailing list Yum@xxxxxxxxxxxxxxxxxxxx https://lists.dulug.duke.edu/mailman/listinfo/yum
