- 03 Jul, 2007 23 commits
-
-
Roy Zang authored
Signed-off-by:
Roy Zang <tie-fei.zang@freescale.com> Signed-off-by:
Kumar Gala <galak@kernel.crashing.org>
-
Roy Zang authored
Document the global utilities node define and example. Signed-off-by:
-
Kazunori Asayama authored
The function backing_ops->read_mfc_tagstatus() doesn't return a correct value because the dma_tagstatus_R register isn't saved in CSA. This fixes the problem. Signed-off-by:
Kazunori Asayama <asayama@sm.sony.co.jp> Signed-off-by:
Jeremy Kerr <jk@ozlabs.org> Signed-off-by:
Paul Mackerras <paulus@samba.org>
-
Kazunori Asayama authored
When waiting for I/O events on mfc in an SPU context by using poll/epoll syscalls, some of the events can be lost because of wrong order of poll_wait and MFC status checks in the spufs_mfc_poll function and non-atomic update of tagwait. This fixes the problem. Signed-off-by:
-
Christoph Hellwig authored
Export spu statistics in sysfs. Signed-off-by:
Christoph Hellwig <hch@lst.de> Signed-off-by:
Arnd Bergmann <arnd.bergmann@de.ibm.com> Signed-off-by:
-
Christoph Hellwig authored
spu_activate can be called from multiple threads at the same time on behalf of the same spu context. We need to make sure to only add it once to avoid runqueue corruption. Signed-off-by:
-
Christoph Hellwig authored
Only enable the scheduler tick if we have any context waiting to be scheduled. Signed-off-by:
-
Jeremy Kerr authored
We're currently too permissive with counting libassist calls - fix the check on the SPE stop-and-signal status. Signed-off-by:
-
Christoph Hellwig authored
Export per-context statistics in spufs. Signed-off-by:
-
Christoph Hellwig authored
Provide load average information for spu context. The format is identical to /proc/loadavg, which is also where a lot of code and concepts is borrowed from. Signed-off-by:
-
Christoph Hellwig authored
The new tid file contains the ID of the thread currently running the context, if any. This is used so that the new spu-top and spu-ps tools can find the thread in /proc. Signed-off-by:
-
Jeremy Kerr authored
Remove redundant whitespace in arch/powerpc/platforms/cell/spufs/ Signed-off-by:
-
Jeremy Kerr authored
spufs_dir_inode_operations is exactly the same as simple_dir_inode_operations. Use that instead. Signed-off-by:
-
Christoph Hellwig authored
And last but not least we need to make sure the scheduler tick never preempts a nosched context. Signed-off-by:
-
Christoph Hellwig authored
spu_deactivate should never be called for nosched contets. Put in a check so we can print a stacktrace and exit early in case it happes erroneously. Signed-off-by:
-
Christoph Hellwig authored
Add a cpus_allowed allowed filed to struct spu_context so that we always use the cpu mask of the owning thread instead of the one happening to call into the scheduler. Also use this information in grab_runnable_context to avoid spurious wakeups. Signed-off-by:
-
Christoph Hellwig authored
Update scheduling information on every spu_run to allow for setting threads to realtime priority just before running them. This requires some slightly ugly code in spufs_run_spu because we can just update the information unlocked if the spu is not runnable, but we need to acquire the active_mutex when it is runnable to protect against find_victim. This locking scheme requires opencoding spu_acquire_runnable in spufs_run_spu which actually is a nice cleanup all by itself. Signed-off-by:
-
Jeremy Kerr authored
Print out a few scheduler tuning parameters when we've compiled with DEBUG defined. Signed-off-by:
-
Jeremy Kerr authored
The current timeslice code mixes 'jiffies' up with 'spesched ticks'. This change correctly defines the number of time slices each SPE contexts is given, and clarifies the comment. This brings the default timeslice for SPE contexts into a reasonable range. Signed-off-by:
-
Christoph Hellwig authored
Enable preemptive scheduling for non-RT contexts. We use the same algorithms as the CPU scheduler to calculate the time slice length, and for now we also use the same timeslice length as the CPU scheduler. This might be not enough for good performance and can be changed after some benchmarking. Note that currently we do not boost the priority for contexts waiting on the runqueue for a long time, so contexts with a higher nice value could starve ones with less priority. This could easily be fixed once the rework of the spu lists that Luke and I discussed is done. Signed-off-by:
-
Christoph Hellwig authored
Get rid of the scheduler workqueues that complicated things a lot to a dedicated spu scheduler thread that gets woken by a traditional scheduler tick. By default this scheduler tick runs a HZ * 10, aka one spu scheduler tick for every 10 cpu ticks. Currently the tick is not disabled when we have less context than available spus, but I will implement this later. Signed-off-by:
-
Sebastian Siewior authored
Add a bit define from book, and replace one hex number with a symbol, for clarity. Signed-off-by:
Sebastian Siewior <bigeasy@linux.vnet.ibm.com> Signed-off-by:
-
Sebastian Siewior authored
Currently it fails with gcc from sdk 2.1 because of a spec change [1]. Maybe we should start using the definitions from spu_mfcio.h. [1] http://gcc.gnu.org/ml/gcc-patches/2006-11/msg01598.htmlSigned-off-by:
-
- 29 Jun, 2007 17 commits
-
-
Kumar Gala authored
Signed-off-by: -
Kumar Gala authored
Use the ppc64 style list management and allocation functions for pci_controllers. This makes the pci_controller structs just a bit more common between ppc32 & ppc64. Signed-off-by: -
Kumar Gala authored
Moved the low hanging fruit that was either identical or close to it between ppc32 & ppc64 for PCI into pci-common.c Signed-off-by: -
Kumar Gala authored
In the places we can move to using pci_bus_to_host, this allows us to make pci_bus_to_host static and remove its export. Signed-off-by: -
Kumar Gala authored
Make the ppc32 pcibios_alloc_controller take a device node to match the ppc64 prototypes and have it set arch_data. Signed-off-by: -
Kumar Gala authored
Make the pci_controller struct use global_number for the PHB domain number instead of index to match what ppc64 does and reuse its pci_domain_nr code. Introduced a pci-common.c to handle shared code between ppc32 & ppc64. Signed-off-by: -
Kumar Gala authored
There are no in kernel users of any off these functions and some of them were not even EXPORT_SYMBOL: - pci_bus_io_base() - pci_bus_io_base_phys() - pci_bus_mem_base_phys() - pci_resource_to_bus() - phys_to_bus() - pci_phys_to_bus() - pci_bus_to_phys() - pci_init_resource() - resource_fixup() Signed-off-by: -
Kumar Gala authored
We never actually read local_number so lets just remove it. Signed-off-by: -
Kumar Gala authored
Moved bits need for ppc32 from asm-ppc/pci-bridge.h into asm-powerpc/pci-bridge.h. Removed ARCH=powerpc specific bits (and comments related to ARCH=ppc) from asm-ppc/pci-bridge.h as its only used on ARCH=ppc. Signed-off-by: -
Zhang Wei authored
The Freescale PCI-e RC poses as a transparent bridge, but does not implement the IO_BASE or IO_LIMIT registers in the config space. This means that the code which initializes the bridge resources ends up setting the IO resources erroneously. Add quick_fsl_pcie_transparent() to handle this. This change sets RC of mpc8641 to be a transparent bridge for legacy I/O access and initializes the RC bridge resources from the device tree. Signed-off-by:
Zhang Wei <wei.zhang@freescale.com> Signed-off-by:
Andy Fleming <afleming@freescale.com> Signed-off-by:
Jon Loeliger <jdl@freescale.com> Signed-off-by:
-
York Sun authored
In pcibios_fixup_bus(), bridges that are subordinate to transparent bridges were still relocating their IORESOURCE_IO and IO_RESOURCE_MEM start and end values. Fix this by preventing the transparent bridge from relocating the start and end values, thus allowing the subordinate non-transparent bridge full molestation rights. Signed-off-by:
York Sun <yorksun@freescale.com> Signed-off-by:
-
Zhang Wei authored
Set IDE in ULI1575 to not 100% native mode, which forces the IDE driver to probe the irq itself. Signed-off-by:
-
Kumar Gala authored
The Freescale PCI-e controllers have an issue in that they use the PCI_PRIMARY_BUS register in the virtual P2P bridge to determine which bus number to match on when generating a type 0 config cycle. The issue is if we are renumbering bus numbers to match Linux we will try setting the PCI_PRIMARY_BUS and will not know which bus number to use for generating type 0 config cycles. We surpress writing the register in the P2P bridge and always keep it at zero. In the future when proper PCI domain support is working we should be able to remove this. Signed-off-by: -
Zhang Wei authored
We check the Link Training and State Status register to make sure we are at least at the L0 state. Signed-off-by:
-
Kumar Gala authored
The generic PCI config ops indirect support for ppc32 covers only two cases (implicit vs explicit) type 0/1 config cycles via set_cfg_type. Added a indirect_type bit mask to handle other variants. Added support for PCI-e extended registers and moved the cfg_type handling into the bit mask for ARCH=powerpc. We can also use this to handle indirect quirks. Signed-off-by: -
Wade Farnsworth authored
This adds device nodes for the PCI bridges as well as the ISA devices on the newer revision MPC8641HPCN. It also adds the PCI ranges to the soc node so that address translation for the ISA devices works properly. Signed-off-by:
Wade Farnsworth <wfarnsworth@mvista.com> Signed-off-by:
-
Zhang Wei authored
Remove errata for PCI-e support of Rev 1.0 of MPC8641 since its considered obselete and is not production level silicon from Freescale. Signed-off-by:
-
