need_active_balance() gates the asymmetric packing based due to power
save logic, but for packing we don't care.
This marks the type of balanace we are attempting to do perform from
f_b_g() and stops need_active_balance() power save logic gating a
balance in the asymmetric packing case.
Signed-of
With the asymmetric packing infrastructure, fix_small_imbalance is
causing idle higher threads to pull tasks off lower threads.
This is being caused by an off-by-one error.
Signed-off-by: Michael Neuling
---
I'm not sure this is the right fix but without it, higher threads pull
tasks off the
The POWER7 core has dynamic SMT mode switching which is controlled by
the hypervisor. There are 3 SMT modes:
SMT1 uses thread 0
SMT2 uses threads 0 & 1
SMT4 uses threads 0, 1, 2 & 3
When in any particular SMT mode, all threads have the same performance
as each other (ie. a
Some CPUs perform better when tasks are run on lower thread numbers.
In the case of POWER7, when higher threads are idled, the core can run
in lower SMT modes and hence perform better.
This creates a new sd flag to prefer lower threads.
Based heavily on patch from Peter Zijlstra.
Signed-off-b
When calculating capacity we use the following calculation:
capacity = DIV_ROUND_CLOSEST(power, SCHED_LOAD_SCALE);
In SMT2, power will be 1178/2 (provided we are not scaling power with
freq say) and SCHED_LOAD_SCALE will be 1024, resulting in capacity
being 1.
With SMT4 however, power wil
This patch series implements asymmetric SMT packing which ensures
consistently good performance on POWER7.
I'm no scheduler expert so these probably need some review. It works
for my simple test case on a POWER7 with SMT4. I can now run a number
of CPU bound processes and they are pulled down
On Apr 8, 2010, at 5:53 AM, Felix Radensky wrote:
> Hi,
>
> I've just noticed that ethernet nodes are missing in p1020rdb device tree.
> Are there any known problems with ethernet controllers on p1020rdb ?
I'd say its not problems, its support for the newer controller on p1020rdb. At
the time
Wolfram Sang wrote:
> Made necessary by 6992f5334995af474c2b58d010d08bc597f0f2fe. Prevents further
> "key xxx not in .data" bug-reports. Although some attributes could probably be
> converted to static ones, this is left for people having hardware to test.
>
> Found by this semantic patch:
>
> @
Howdy All,
I have (almost) successfully upgraded from Linux 2.6.22 to 2.6.26 (both
downloaded from debian) on my mpc8347 powerpc, but I think I may be
missing a required change to my dts regarding the IPIC or maybe a change
in how my driver requests IRQs.
I have several modules that are maintaine
> > When a DTLB Miss exception can't find a PTE for the virtual address
being
> > written/read, it dummies up the SPRs for a DSI exception and then
calls
> > directly into the DSI exception code. Just before the DTLB miss code
> > stores a value into DSISR it sets bit 2, which for a DSI excep
This patch adds support for using the LL TEMAC Ethernet driver on
non-Virtex 5 platforms by adding support for accessing the Soft DMA
registers as if they were memory mapped instead of solely through the
DCR's (available on the Virtex 5).
The patch also updates the driver so that it runs on the Mi
The virt_to_bus call should not be used any longer as it's
considered illegal. The driver has the physical address of
the buffer in the descriptor such that it's not necessary
anyway.
Signed-off-by: John Linn
---
drivers/net/ll_temac_main.c |4 +---
1 files changed, 1 insertions(+), 3 deleti
On Mon, 2010-03-29 at 23:24 -0700, bruce_leon...@selinc.com wrote:
> I'm tracking a problem that's leading me through DSI and DTLB miss
> exceptions on an MPC8347 (e300c1 core), and I've come across an oddity
> that I'm hoping someone can explain.
>
> When a DTLB Miss exception can't find a PTE
Hi,
I've just noticed that ethernet nodes are missing in p1020rdb device tree.
Are there any known problems with ethernet controllers on p1020rdb ?
Thanks.
Felix.
___
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/lis
Hello,
I found out this strange behavior on my system.
I’m working on an embedded device with a linux kernel 2.6.24.6, we’re
using a modified Ångström distribution with BusyBox v1.13.2
I’m having problem when I try to copy files larger than 52 Megabytes.
The system hangs and I have a kernel pa
Hi List,
I am using linux 2.6.22 on a MPC8358E custom board.
I use a simple userspace process that periodically (100 ms) calls an ioctl
to a watchdog driver to toggle its pin state.
The problem is that when I insert or remove the Ethernet cable, this process
is not scheduled for execution. The w
Hi ,
We are using gianfar ethernet driver in our device (2.6.20 kernel). I
have observed the following error after i run high throughput traffic
for some time.
NETDEV WATCHDOG: eth0: transmit timed out
The Kernel we are using is 2.6.20 kernel. I have added the work queue
support in the gfar_tim
17 matches
Mail list logo