| To: | linux-xfs@xxxxxxxxxxx |
|---|---|
| Subject: | TAKE - xfs_dm_send_data_event() breaks some HSMs |
| From: | Dean Roehrich <roehrich@xxxxxxx> |
| Date: | Thu, 22 Aug 2002 15:04:59 -0500 (CDT) |
| Sender: | owner-linux-xfs@xxxxxxxxxxx |
Date: Thu Aug 22 13:04:46 PDT 2002
Workarea: clink-eth.americas.sgi.com:/data/clink/a67/roehrich/2.4.x-xfs
The following file(s) were checked into:
bonnie.engr.sgi.com:/isms/slinx/2.4.x-xfs
Modid: 2.4.x-xfs:slinx:125722a
linux/fs/xfs/xfs_dmapi.c - 1.76
- We'll use dmstate to keep a counter and use this to coordinate
activity between xfs_dm_punch_hole and xfs_dm_send_data_event.
Checking DM_EVENT_ENABLED got us into trouble. By checking that
event,
after you've already sent it, you're dictating behavior to the
HSM--that it
must clear the event's bit from the event mask before responding to
the event.
The DMAPI spec does not require that HSMs behave this way, and some,
like
Veritas's, don't.
|
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: Aligning v2 logs on LVM2/device-mapper volumes, Steve Lord |
|---|---|
| Next by Date: | TAKE - remove some debug code in last dmapi mod, Dean Roehrich |
| Previous by Thread: | Aligning v2 logs on LVM2/device-mapper volumes, ERG |
| Next by Thread: | TAKE - remove some debug code in last dmapi mod, Dean Roehrich |
| Indexes: | [Date] [Thread] [Top] [All Lists] |