On 01/23/2018 05:28 PM, Brian Foster wrote:
On Tue, Jan 23, 2018 at 04:57:03PM +0200, Avi Kivity wrote:
I'm seeing the equivalent[*] of xfs_extent_busy_flush() sleeping in my
beautiful io_submit() calls.
Questions:
- Is it correct that RWF_NOWAIT will not detect the condition that led to
the log being forced?
- If so, can it be fixed?
- Can I do something to reduce the odds of this occurring? larger logs,
more logs, flush more often, resurrect extinct species and sacrifice them to
the xfs gods?
- Can an xfs developer do something? For example, make it RWF_NOWAIT
friendly (if the answer to the first question was "correct")
So RWF_NOWAIT eventually works its way to IOMAP_NOWAIT, which looks like
it skips any write call that would require allocation in
xfs_file_iomap_begin(). The busy flush should only happen in the block
allocation path, so something is missing here. Do you have a backtrace
for the log force you're seeing?
Here's a trace. It's from a kernel that lacks RWF_NOWAIT.
0xffffffff816ab231 : __schedule+0x531/0x9b0 [kernel]
0xffffffff816ab6d9 : schedule+0x29/0x70 [kernel]
0xffffffff816a90e9 : schedule_timeout+0x239/0x2c0 [kernel]
0xffffffff816aba8d : wait_for_completion+0xfd/0x140 [kernel]
0xffffffff810ab41d : flush_work+0xfd/0x190 [kernel]
0xffffffffc00ddb3a : xlog_cil_force_lsn+0x8a/0x210 [xfs]
0xffffffffc00dbbf5 : _xfs_log_force+0x85/0x2c0 [xfs]
0xffffffffc00dbe5c : xfs_log_force+0x2c/0x70 [xfs]
0xffffffffc0078f60 : xfs_alloc_ag_vextent_size+0x250/0x630 [xfs]
0xffffffffc0079ed5 : xfs_alloc_ag_vextent+0xe5/0x150 [xfs]
0xffffffffc007abc6 : xfs_alloc_vextent+0x446/0x5f0 [xfs]
0xffffffffc008b123 : xfs_bmap_btalloc+0x3f3/0x780 [xfs]
0xffffffffc008b4be : xfs_bmap_alloc+0xe/0x10 [xfs]
0xffffffffc008bef9 : xfs_bmapi_write+0x499/0xab0 [xfs]
0xffffffffc00c6ec8 : xfs_iomap_write_direct+0x1b8/0x390 [xfs]
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html