If GetNameOwner fails with an error that is not NameHasNoOwner, dbus_g_proxy_manager_unregister wrongly asserts that the proxy is in a list to which it was never added. (Based on a patch from Janne Karhunen, in Maemo's dbus-glib; his patch warned in this situation, but if GetNameOwner failed with some other error, then we already warned about that anyway.)
Created attachment 48101 [details] [review] dbus_g_proxy_manager_unregister: if GetNameOwner failed, don't assert got_name_owner_cb never adds the proxy to the unassociated_proxies list if there is a D-Bus error other than NameHasNoOwner. Based on a patch from Janne Karhunen, for Maemo. Bug-NB: NB#116862
Without any reviewer hat: pseudo-review++ No reason to assert, also there are several other occurences of g_assert which would deserve a review.
(In reply to comment #2) > Without any reviewer hat: pseudo-review++ Thanks, fixed in git for 0.96 > No reason to assert, also there are several other occurences of g_assert which > would deserve a review. Care to name some? (Or review them?) My branches for Bug #35766 and Bug #35767 might well fix some wrong assertions too.
(In reply to comment #3) > Care to name some? (Or review them?) I'm giving a look at them and I'll open another bug if I'll found out anything suspicious, they are probably unrelated to this one. > My branches for Bug #35766 and Bug #35767 might well fix some wrong assertions > too. Cool, I'll check against those to avoid duplicate work, thanks
(In reply to comment #3) > Care to name some? (Or review them?) False alarm, I thought I saw an assertion used in place of a condition, instead of a consistency check. Didn't find again, I probably misread the block.
*** Bug 41339 has been marked as a duplicate of this bug. ***
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.