Re: Crash in pgCrypto?
От | Andrew Dunstan |
---|---|
Тема | Re: Crash in pgCrypto? |
Дата | |
Msg-id | 4856E281.8000609@dunslane.net обсуждение исходный текст |
Ответ на | Re: Crash in pgCrypto? (David Fetter <david@fetter.org>) |
Ответы |
Re: Crash in pgCrypto?
|
Список | pgsql-hackers |
David Fetter wrote: > On Mon, Jun 16, 2008 at 02:48:42PM -0400, Robert Treat wrote: > >> On Monday 16 June 2008 09:54:16 Alvaro Herrera wrote: >> >>> Mario Weilguni wrote: >>> >>>> Could someone using the pgcrypto extension please verify this? >>>> >>>> SELECT encode(digest(null, 'md5'::text), 'hex'); >>>> or >>>> SELECT digest(null, 'md5'); >>>> >>>> Takes a few seconds, and then crashes the server with a Signal 11. My >>>> system is PostgreSQL 8.2.7. Seems to be an unchecked access to memory >>>> location 0. >>>> >>> I think the functions were made STRICT recently, and the NULL checks >>> were removed, but people with the old definitions of the functions could >>> see the crashes. Try removing pgcrypto and recompiling it from a fresh >>> release. >>> >> <broken record> >> I still advocate to folks to try to put contrib modules into thier >> own schemas whenever possible, so that you have the option of doing >> pg_dump -N contribmodule, makes things like this much easier to work >> around. (And yes, I've volunteered to patch the contribs with this >> if we ever decide to make it the default setup) >> </broken record> >> > > I, too, would be happy to do the legwork on this one. I believe we'd > want to have both per-db and per-role settings for search_path. > What's involved with creating that latter? > > > Proper support for module install / uninstall will be a far better solution. Why would you wast your time on something that will be at best half-baked? cheers andrew
В списке pgsql-hackers по дате отправления: