On Thu, 2007-07-05 at 16:55 -0500, Scott Wood wrote:
> As revealed by the recent "Prevent data exception in kernel space"
> patch, versions of glibc prior to 2.4[1] assume that, on powerpc32, they
> can execute out of any readable mapping, regardless of whether it is
> marked for execution. Thi
Please add to patch header:
Newer asm-generic/vmlinux.lds.h provides a DATA_DATA macro which defines
the content of what is normally referred to by *(.data). It permits
adding more "subsections" (declared by gcc with the section attribute or
in assembly), whichwill be linked in the .data section.
As revealed by the recent "Prevent data exception in kernel space"
patch, versions of glibc prior to 2.4[1] assume that, on powerpc32, they
can execute out of any readable mapping, regardless of whether it is
marked for execution. This happens in the elf_machine_load_address()
function.
To ma
The patch titled
powerpc: add missing DATA_DATA
has been added to the -mm tree. Its filename is
add-missing-data_data-in-powerpc.patch
*** Remember to use Documentation/SubmitChecklist when testing your code ***
See http://www.zip.com.au/~akpm/linux/patches/stuff/added-to-mm.txt to fi
>> There are no dma_mmap_x() functions in arch/powerpc. Can someone
>> tell me what the powerpc
>> equivalent to dma_mmap_writecombine() is?
>
> Not sure exactly what arm does here, but it sounds like you want
> to call remap_pfn_range with the _PAGE_NO_CACHE bit set in the
> protection flags, a
Johannes Berg wrote:
> This one doesn't break 32-bit build by simply disabling irqtrace for
> 32-bit.
> --- linux-2.6-git32.orig/include/asm-powerpc/hw_irq.h 2007-06-28
> 14:53:58.730430447 +0200
> +++ linux-2.6-git32/include/asm-powerpc/hw_irq.h 2007-06-28
> 14:58:41.508430447 +0200
> @@ -
On Thu, Jul 05, 2007 at 11:28:38AM -0700, Andrew Morton wrote:
> Well you've sent it a couple of times, and I've sent it in five more times
> over the past year. Once we were told "awaiting maintainer ack".
>
> This situation is fairly stupid. How about we make you the maintainer?
Last time I l
On Thu, Jul 05, 2007 at 08:50:21PM +0200, Gabriel Paubert wrote:
> On Thu, Jul 05, 2007 at 12:03:00PM -0500, Olof Johansson wrote:
> > Workaround for errata 4290: The decrementer ticks at half the expected rate
> > so load it with half the value it would otherwise be loaded with.
> >
> >
> > Inde
On Thu, Jul 05, 2007 at 12:03:00PM -0500, Olof Johansson wrote:
> Workaround for errata 4290: The decrementer ticks at half the expected rate
> so load it with half the value it would otherwise be loaded with.
>
>
> Index: 2.6.21/include/asm-powerpc/time.h
> ==
On Mon, 2 Jul 2007 13:39:17 -0500
[EMAIL PROTECTED] (Linas Vepstas) wrote:
>
> Various PCI bus errors can be signaled by newer PCI controllers.
> This patch adds the PCI error recovery callbacks to the Symbios
> SCSI device driver. The patch has been tested, and appears to
> work well.
>
y
On Thursday 05 July 2007, Timur Tabi wrote:
> static int at91_pcm_mmap(struct snd_pcm_substream *substream,
> struct vm_area_struct *vma)
> {
> struct snd_pcm_runtime *runtime = substream->runtime;
>
> return dma_mmap_writecombine(substream->pcm->card->dev, vma,
>
Workaround for Errata 4504: PMC0 doesn't seem to set SIARLOG when
an exception occurs. Instead, log the PC for when we took the actual
exception.
Index: 2.6.21/arch/powerpc/oprofile/op_model_pa6t.c
===
--- 2.6.21.orig/arch/powerpc/o
Our PURR ticks at the wrong rate (same as the decrementer). Since it
doesn't really have any value-add besides on partitioned and multithreaded
systems, refrain from using it at the moment.
Index: 2.6.21/include/asm-powerpc/cputable.h
==
Don't allow unguarded mappings in PCI-express space. This is resolves
problems with frame buffers having garbage on them due to failing
write merges.
Index: 2.6.21/arch/powerpc/mm/hash_native_64.c
===
--- 2.6.21.orig/arch/powerpc/mm/
Workaround for errata 4713: Concurrent access to some of the devices
in the SOC will result in a bus hang.
To avoid this, we introduce the lbi_lock, and grab it in the drivers
in question. Not a pretty solution, but adequate for now.
Index: mainline/arch/powerpc/platforms/pasemi/setup.c
Always keep a page mapped at EA 0. Right now it is shared among all
processes, as long as noone does a NULL pointer deref and uses the data
there, there should be no security implications. Still, this is better
not used in any kind of environment where users can't be trusted.
Index: mainline/arch
Driver for the CompactFlash slot on the PA Semi Electra eval board. It's
a simple device sitting on localbus, with interrupts and detect/voltage
control over GPIO.
The driver is implemented as an of_platform driver, and adds localbus
as a bus being probed by the of_platform framework.
Signed-off
Use MSR_PMM to disable performance monitor counting while in idle,
instead of using the CTRL register like on the IBM implementation.
This needs further cleanup. It should be possible to switch over the
default behaviour to MSR_PMM on IBM hardware too.
Index: 2.6.21/arch/powerpc/kernel/process.c
Errata 4025 workaround: A SLB load can in some cases be corrupted. As a
workaround,
insert an instruction with sync side effects after the load (in this case it's
the
mtmsrd).
Index: 2.6.21/arch/powerpc/kernel/entry_64.S
===
--- 2.6
Errata 4628 causes lockups during table walks that are recovered from
when a ptesync is received. So, issue a ptesync every timer interrupt,
and for CPUs that have not yet spun up, at regular intervals. This seems
sufficient to avoid hitting it.
Index: 2.6.21/arch/powerpc/kernel/head_64.S
Errata 4161: Concurring ptesyncs on both cores might lock up the
cores. Most ptesyncs are already under tlbie_lock, so use that in the only
other remaining location (the timer interrupt ones we've added ourselves).
Index: 2.6.21/arch/powerpc/kernel/time.c
=
Workaround for errata 4111: IPI registers of the OpenPIC don't read back
right.
Also, reading back regular irq setup registers is broken, so keep a
software copy.
Index: mainline/arch/powerpc/sysdev/mpic.c
===
--- mainline.orig/arch
Support for, and workarounds for various problems with our UART.
This needs further cleanup at some point.
Index: mainline/arch/powerpc/kernel/udbg_16550.c
===
--- mainline.orig/arch/powerpc/kernel/udbg_16550.c
+++ mainline/arch/pow
Workaround for errata 4290: The decrementer ticks at half the expected rate
so load it with half the value it would otherwise be loaded with.
Index: 2.6.21/include/asm-powerpc/time.h
===
--- 2.6.21.orig/include/asm-powerpc/time.h
+++
The name says it all. Try to provide more good info on machine checks.
Index: 2.6.21/arch/powerpc/platforms/pasemi/setup.c
===
--- 2.6.21.orig/arch/powerpc/platforms/pasemi/setup.c
+++ 2.6.21/arch/powerpc/platforms/pasemi/setup.c
@@
Workaround for errata 4712; on an SMP system the PIC can get stuck
trying to read the next available interrupt from INTACK in case there's
only a spurious available. The workaround is to program timers to cause
interrupts every now and then, which recovers from the lockup.
Index: mainline/arch/po
Remove idle_spin function pointer, since it just seems to make the idle
loop run hotter by doing bl/blr.
Index: 2.6.21/arch/powerpc/platforms/pasemi/idle.c
===
--- 2.6.21.orig/arch/powerpc/platforms/pasemi/idle.c
+++ 2.6.21/arch/power
Abstract out the PCI config read/write accesses into reg read/write ones, still
calling the pci accessors on the back end.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: netdev-2.6/drivers/net/pasemi_mac.c
===
--- netdev-2
Export some of the HID registers via sysfs. Useful when debugging.
Index: 2.6.21/arch/powerpc/kernel/sysfs.c
===
--- 2.6.21.orig/arch/powerpc/kernel/sysfs.c
+++ 2.6.21/arch/powerpc/kernel/sysfs.c
@@ -197,6 +197,17 @@ SYSFS_PMCSETUP(p
Add 88E1112 PHY ID to the marvell driver.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: mainline/drivers/net/phy/marvell.c
===
--- mainline.orig/drivers/net/phy/marvell.c
+++ mainline/drivers/net/phy/marvell.c
@@ -252,6 +
Move away from using the pci config access functions for simple register
access. Our device has all of the registers in the config space (hey,
from the hardware point of view it looks reasonable :-), so we need to
somehow get to it. Newer firmwares have it in the device tree such that
we can just
Enable settings to target L2 for the first few cachelines of the packet,
since we'll access them to get to the various headers.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: netdev-2.6/drivers/net/pasemi_mac.c
===
--- ne
Postpone pci unmap and skb free of the transmitted buffers to outside
of the tx ring lock, batching them up 32 at a time.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: netdev-2.6/drivers/net/pasemi_mac.c
===
--- netdev-2.
Various minor performance tweaks, do some explicit prefetching of packet
data, etc.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: netdev-2.6/drivers/net/pasemi_mac.c
===
--- netdev-2.6.orig/drivers/net/pasemi_mac.c
+++ net
Enable LLTX on pasemi_mac: we're already doing sufficient locking
in the driver to enable it.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: netdev-2.6/drivers/net/pasemi_mac.c
===
--- netdev-2.6.orig/drivers/net/pasemi_ma
It was mistakenly set to interrupt on the second packet instead of first,
causing
some interesting latency behaviour.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: netdev-2.6/drivers/net/pasemi_mac.c
===
--- netdev-2.6.o
The following series is a snapshot of the in-progress and out-of-tree
errata workarounds that I'm currently maintaining.
The first patches are mostly copies of stuff that has already been sent
out for upstream merge, followed by some patches that are not yet ready
to be merged (need cleanup, more
Clean the TX ring in the poll call, to avoid sitting on mapped buffers
for a long time. NFS doesn't seem to like it much, for example.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: netdev-2.6/drivers/net/pasemi_mac.c
===
EDAC driver for the memory controllers on PA Semi PA6T-1682M.
Signed-off-by: Egor Martovetsky <[EMAIL PROTECTED]>
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: 2.6.21/drivers/edac/pasemi_edac.c
===
--- /dev/null
+++ 2.6.21
ppc64 really needs ioaddr_t to be 32-bit, since I/O addresses really are
MMIO addresses, and remapped at an offset that's well above 16 bits in
some cases.
While the type is exported to userspace, there hasn't been any platforms
with PCMCIA on 64-bit powerpc until now, so changing it won't regress
I'm porting a driver from ARM to PowerPC, and I came across this function:
static int at91_pcm_mmap(struct snd_pcm_substream *substream,
struct vm_area_struct *vma)
{
struct snd_pcm_runtime *runtime = substream->runtime;
return dma_mmap_writecombine(substream->pcm->card->d
On Tue, 2007-07-03 at 04:17, Kumar Gala wrote:
> I'm still working through the back log of patches related to ppc32/
> fsl. If there are patches people want to make sure get into 2.6.23
> let me know.
> Jon Loeliger (1):
>[POWERPC] Replace use of GET_64BIT(prop, i) with of_read_number(
On Thu, Jul 05, 2007 at 04:24:46PM +0200, Jan-Bernd Themann ([EMAIL PROTECTED])
wrote:
> > I've couple of comments on the driver, but mainly the fact of decreased
> > CPU usage itself - what was the magnitude of the win with this driver,
> > it looks like because of per-packet receive code path in
Driver for the CompactFlash slot on the PA Semi Electra eval board. It's
a simple device sitting on localbus, with interrupts and detect/voltage
control over GPIO.
The driver is implemented as an of_platform driver, and adds localbus
as a bus being probed by the of_platform framework.
Signed-off
Rename the pasemi platform to "pasemi" to be in line with the platforms
directory name.
Signed-off-by: Olof Johansson <[EMAIL PROTECTED]>
Index: mainline/arch/powerpc/platforms/pasemi/setup.c
===
--- mainline.orig/arch/powerpc/platfo
On Wed, Jun 27, 2007 at 06:20:38AM -0500, Milton Miller wrote:
> If CONFIG_PCMCIA=m then your notifier is not registered. The modprobe
> of your of_driver loads ds.ko, registers the bus, then registers your
> driver. When the socket driver tries to dma, the BUG in dma_64 for no
> archdata.dma
On Wed, Jun 27, 2007 at 06:20:27AM -0500, Milton Miller wrote:
> get_property is now a #define for of_get_property. I think the use of
> get_property is being deprecated.
Yeah, it's a pita. I've moved these over now.
> If of_get_property fails to find the property, then you will
> dereference
On Sun, Jul 01, 2007 at 08:49:37PM -0400, Sonny Rao wrote:
> The pseries platform does not have a default function for init_IRQ and
> does not install one if it doesn't find or doesn't recognize an
> interrupt controller in the device tree. Currently, the kernel dies
> when it tries to call the NU
With !CONFIG_NUMA, these are static inlines in the header file so don't
generate exports for them then.
Signed-off-by: Johannes Berg <[EMAIL PROTECTED]>
---
arch/powerpc/kernel/sysfs.c |5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
--- wireless-dev.orig/arch/powerpc/kernel/sysfs.
On Sat, 2007-06-30 at 10:58 +0200, Christoph Hellwig wrote:
> On Thu, Jun 28, 2007 at 06:20:42PM +0200, Johannes Berg wrote:
> > This one doesn't break 32-bit build by simply disabling irqtrace for
> > 32-bit.
>
> This looks really cool to me. I'd love to see this in 2.6.23.
Hmm. Looks like I ju
This patch enables the receive side processing to aggregate TCP packets within
the HEA device driver. It analyses the packets already received after an
interrupt arrived and forwards these as chains of SKBs for the same TCP
connection with modified header field. We have seen a lower CPU load and
im
This patch introduces a capability flag that is used by the DLPAR userspace
tool to check which DLPAR features are supported by the eHEA driver.
Missing goto has been included.
Signed-off-by: Jan-Bernd Themann <[EMAIL PROTECTED]>
---
drivers/net/ehea/ehea.h |8 +++-
drivers/net/ehea
52 matches
Mail list logo
