Re: db bug with i386: sanity check
Re: db bug with i386: sanity check
- Subject: Re: db bug with i386: sanity check
- From: Stephane Sudre <email@hidden>
- Date: Wed, 22 Mar 2006 16:03:13 +0100
On 21 mars 06, at 19:17, Eric Albert wrote:
On Mar 21, 2006, at 5:35 AM, Stephane Sudre wrote:
The more I look into this db thing, the more I believe there is a bug
in the db code.
So before filing a bug report, I may need a sanity check and hope
there's a doctor in the room.
It sounds like a bug to me. Go ahead and write it up. Thanks!
Done. Actually I probably found where the problem exactly lies.
It's not with the put API, it's with the sync API.
If you do something like:
open
get
put
sync
exit(0);
It works fine on an any platform.
If you do something like:
open
get
put
sync
sync
exit(0);
It corrupts the data base on the Intel platform (PPC is fine). Doing 2
or more sync in the same execution is killing the data base. At least
that's what I'm observing.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Darwin-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden