Revision ae1e206b806ccc490dadff59af8a7a2477b32884 authored by Josef Bacik on 07 August 2012, 20:00:32 UTC, committed by Chris Mason on 28 August 2012, 20:53:38 UTC
Daniel Blueman reported a bug with fio+balance on a ramdisk setup. Basically what happens is the balance relocates a tree block which will drop the implicit refs for all of its children and adds a full backref. Once the block is relocated we have to add the implicit refs back, so when we cow the block again we add the implicit refs for its children back. The problem comes when the original drop ref doesn't get run before we add the implicit refs back. The delayed ref stuff will specifically prefer ADD operations over DROP to keep us from freeing up an extent that will have references to it, so we try to add the implicit ref before it is actually removed and we panic. This worked fine before because the add would have just canceled the drop out and we would have been fine. But the backref walking work needs to be able to freeze the delayed ref stuff in time so we have this ever increasing sequence number that gets attached to all new delayed ref updates which makes us not merge refs and we run into this issue. So to fix this we need to merge delayed refs. So everytime we run a clustered ref we need to try and merge all of its delayed refs. The backref walking stuff locks the delayed ref head before processing, so if we have it locked we are safe to merge any refs inside of the sequence number. If there is no sequence number we can merge all refs. Doing this not only fixes our bug but keeps the delayed ref code from adding and removing useless refs and batching together multiple refs into one search instead of one search per delayed ref, which will really help our commit times. I ran this with Daniels test and 276 and I haven't seen any problems. Thanks, Reported-by: Daniel J Blueman <daniel@quora.org> Signed-off-by: Josef Bacik <jbacik@fusionio.com>
1 parent 5a24e84
clock.txt
Clock scaling
-------------
The kernel supports scaling of CLCK.CMODE, CLCK.CM and CLKC.P0 clock
registers. If built with CONFIG_PM and CONFIG_SYSCTL options enabled, four
extra files will appear in the directory /proc/sys/pm/. Reading these files
will show:
p0 -- current value of the P0 bit in CLKC register.
cm -- current value of the CM bits in CLKC register.
cmode -- current value of the CMODE bits in CLKC register.
On all boards, the 'p0' file should also be writable, and either '1' or '0'
can be rewritten, to set or clear the CLKC_P0 bit respectively, hence
controlling whether the resource bus rate clock is halved.
The 'cm' file should also be available on all boards. '0' can be written to it
to shift the board into High-Speed mode (normal), and '1' can be written to
shift the board into Medium-Speed mode. Selecting Low-Speed mode is not
supported by this interface, even though some CPUs do support it.
On the boards with FR405 CPU (i.e. CB60 and CB70), the 'cmode' file is also
writable, allowing the CPU core speed (and other clock speeds) to be
controlled from userspace.
Determining current and possible settings
-----------------------------------------
The current state and the available masks can be found in /proc/cpuinfo. For
example, on the CB70:
# cat /proc/cpuinfo
CPU-Series: fr400
CPU-Core: fr405, gr0-31, BE, CCCR
CPU: mb93405
MMU: Prot
FP-Media: fr0-31, Media
System: mb93091-cb70, mb93090-mb00
PM-Controls: cmode=0xd31f, cm=0x3, p0=0x3, suspend=0x9
PM-Status: cmode=3, cm=0, p0=0
Clock-In: 50.00 MHz
Clock-Core: 300.00 MHz
Clock-SDRAM: 100.00 MHz
Clock-CBus: 100.00 MHz
Clock-Res: 50.00 MHz
Clock-Ext: 50.00 MHz
Clock-DSU: 25.00 MHz
BogoMips: 300.00
And on the PDK, the PM lines look like the following:
PM-Controls: cm=0x3, p0=0x3, suspend=0x9
PM-Status: cmode=9, cm=0, p0=0
The PM-Controls line, if present, will indicate which /proc/sys/pm files can
be set to what values. The specification values are bitmasks; so, for example,
"suspend=0x9" indicates that 0 and 3 can be written validly to
/proc/sys/pm/suspend.
The PM-Controls line will only be present if CONFIG_PM is configured to Y.
The PM-Status line indicates which clock controls are set to which value. If
the file can be read, then the suspend value must be 0, and so that's not
included.
Computing file changes ...
