than 12 queues.
So this series tries to address those issues by switching to use flex array. All
entries were preallocated, and since flex array always do a order-0 allocation,
we can safely increase the limit after.
Only compile test, comments or review are more than welcomed.
Jason Wang (3):
On 06/19/2013 02:31 PM, Eric Dumazet wrote:
> On Wed, 2013-06-19 at 13:40 +0800, Jason Wang wrote:
>> Currently, we use kcalloc to allocate rx/tx queues for a net device which
>> could
>> be easily lead to a high order memory allocation request when initializing a
>> mu
On 06/19/2013 02:34 PM, Eric Dumazet wrote:
> On Wed, 2013-06-19 at 13:40 +0800, Jason Wang wrote:
>> Since we've reduce the size of tun_struct and use flex array to allocate
>> netdev
>> queues, it's safe for us to increase the limit of queues in tuntap.
> Its a
This patches add a read barriers to force the driver to check the interrupt mask
before read_index. Otherwise we may lost a kick to host.
Cc: K. Y. Srinivasan
Cc: Haiyang Zhang
Signed-off-by: Jason Wang
---
drivers/hv/ring_buffer.c |2 ++
1 files changed, 2 insertions(+), 0 deletions
On 06/19/2013 05:56 PM, Eric Dumazet wrote:
> On Wed, 2013-06-19 at 12:11 +0300, Michael S. Tsirkin wrote:
>
>> Well KVM supports up to 160 VCPUs on x86.
>>
>> Creating a queue per CPU is very reasonable, and
>> assuming cache line size of 64 bytes, netdev_queue seems to be 320
>> bytes, that's 320
ip kernel with no patches. When using a 3.10-rc2 tip
kernel with just patches 1-7, the performance improvement of the
workload over the vanilla 3.10-rc2 tip kernel was about 25%.
Tested-by: Jason Low
Thanks,
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel&quo
gt; >>all through the corresponding ARM branch.
> >>
> >Sorry for the tardy response. This cleanup looks really good. If it's
> >not too late:
> >
> >Acked-by: Mike Turquette
> >
> >If the branch is already pulled into arm-soc don't worry
a phy handled by (also modular) mvmdio. The mvmdio
> conversion has been done in
>
> commit c3a07134e6aa5b93a37f72ffa3d11fadf72bf757
> ("mv643xx_eth: convert to use the Marvell Orion MDIO driver")
>
> and should go back any -stable version with that commit (propably only 3.9
On Thu, May 30, 2013 at 11:06:08AM +0200, Arnaud Ebalard wrote:
> Hi Jason and Sebastian,
>
> Sebastian Hesselbarth writes:
>
> > With DT support for mv643xx_eth board specific init for some boards now
> > is unneccessary. Remove those board files, Kconfig entries, and
&
size = ((phys_addr_t)~0);
A similar printk as arm_add_memory for this one too?
printk(KERN_CRIT "Truncating memory at 0x%08llx to fit in "
"32-bit physical address space\n", (long long)start);
Regards,
Jason
--
To unsubscribe from this li
patch 7/8 ( v1 says the linked list were used which is
wrong )
- Other minor bug fixes
Jason Wang (8):
macvtap: do not add self to waitqueue if doing a nonblock read
macvlan: switch to use IS_ENABLED()
macvtap: introduce macvtap_get_vlan()
macvlan: change the max number of queues to 16
Macvtap should be at least compatible with tap, so change the max number to 16.
Signed-off-by: Jason Wang
---
include/linux/if_macvlan.h |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
index e47ad46..62d8bda
Acked-by: Michael S. Tsirkin
Signed-off-by: Jason Wang
---
include/linux/if_macvlan.h |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/include/linux/if_macvlan.h b/include/linux/if_macvlan.h
index 84dde1d..e47ad46 100644
--- a/include/linux/if_macvlan.h
+++ b/include
taps
- [numvtaps, numvtaps + numdisabled) : disabled taps
- [numvtaps + numdisabled, MAX_MAXVTAP_QUEUES) : unused slots
When a tap were enabled and disabled, it was moved to another area.
Signed-off-by: Jason Wang
---
drivers/net/macvtap.c | 167
Factor out the device holding logic to a macvtap_get_vlan(), this will be also
used by multiqueue API.
Signed-off-by: Jason Wang
---
drivers/net/macvtap.c | 27 ---
1 files changed, 20 insertions(+), 7 deletions(-)
diff --git a/drivers/net/macvtap.c b/drivers/net
There's no need to add self to waitqueue if doing a nonblock read. This could
help to avoid the spinlock contention.
Signed-off-by: Jason Wang
---
drivers/net/macvtap.c |7 +--
1 files changed, 5 insertions(+), 2 deletions(-)
diff --git a/drivers/net/macvtap.c b/drivers/net/macv
ned-off-by: Jason Wang
---
drivers/net/macvtap.c | 63 +++-
1 files changed, 20 insertions(+), 43 deletions(-)
diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 56b38e9..e76484c 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.
Signed-off-by: Jason Wang
---
drivers/net/macvtap.c |7 ++-
1 files changed, 2 insertions(+), 5 deletions(-)
diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 03b781c..fac9dbf 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -31,10 +31,6
Though the queue were in fact created by open(), we still need to add this check
to be compatible with tuntap which can let mgmt software use a single API to
manage queues. This patch only validates the device name and moves the TUNSETIFF
to a helper.
Signed-off-by: Jason Wang
---
drivers/net
Arnaud,
On Fri, May 31, 2013 at 12:28:56AM +0200, Arnaud Ebalard wrote:
> Hi,
>
> Jason Cooper writes:
>
> >> For instance 6bd98481ab34 (arm: kirkwood: NETGEAR ReadyNAS Duo v2 init
> >> PCIe via DT) currently sitting in jcooper/mvebu/pcie_kirkwood removes
>
Add check for error pointers returned from get_node_page in order to
avoid dereferencing a bad address on the next use.
Signed-off-by: Jason Hrycay
---
fs/f2fs/xattr.c | 4
1 file changed, 4 insertions(+)
diff --git a/fs/f2fs/xattr.c b/fs/f2fs/xattr.c
index 0b02dce..ae61f35 100644
--- a
On 05/31/2013 10:48 PM, Sergei Shtylyov wrote:
> Hello.
>
> On 31-05-2013 13:53, Jason Wang wrote:
>
>> Though the queue were in fact created by open(), we still need to add
>> this check
>> to be compatible with tuntap which can let mgmt software use a single
&g
On 06/02/2013 07:22 PM, Michael S. Tsirkin wrote:
> On Fri, May 31, 2013 at 05:53:24PM +0800, Jason Wang wrote:
>> This patch adds TUNSETQUEUE ioctl to let userspace can temporarily disable or
>> enable a queue of macvtap. This is used to be compatible at API layer of
>> tunt
4/0x160
> [] register_virtio_driver+0x20/0x40
>
> v2: change msix_vectors uncoditionaly in vp_free_vectors
>
> Cc: Rusty Russell
> Cc: "Michael S. Tsirkin"
> Cc: Jason Wang
> Signed-off-by: Andrew Vagin
> Signed-off-by: Andrey Vagin
> ---
> driver
&vq->mutex);
> vhost_zerocopy_signal_used(n, vq);
> mutex_unlock(&vq->mutex);
> @@ -1091,7 +1096,7 @@ err_used:
> vq->private_data = oldsock;
> vhost_net_enable_vq(n, vq);
> if (ubufs)
> - vhost_net_ubuf_put_an
nusW that you both had patches
depending on (now called) mvebu/of_pci. So we got it into arm-soc
early so those branches could depend on it.
hth,
Jason.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More m
y: Michael S. Tsirkin
> ---
>
> Same as the tun patch really. Also stable material.
Acked-by: Jason Wang
>
> drivers/net/macvtap.c | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
> index 59e96
y: Michael S. Tsirkin
> ---
>
> I haven't figured out why do we get failures,
> but recovery is clearly wrong.
>
> This is also -stable material.
Acked-by: Jason Wang
> drivers/net/tun.c | 5 +++--
> 1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff --git
On Fri, Jun 07, 2013 at 01:59:43PM +0200, Arnd Bergmann wrote:
> On Friday 07 June 2013 18:19:40 Jingoo Han wrote:
> > Hi Jason Gunthorpe,
> >
> > I implemented 'Single domain' with Exynos PCIe for last two months;
> > however, it cannot work properly due to t
that the i2c controller have the
> + Transaction Generator support and we want to use it. Not all the
> + mv64xxx controller have this feature.
Do you have a list of which controllers definitely do, or definitely
don't? That would be helpful for folks adding new boards.
> +
>
and to John and Thomas. I take the patches for
> >> clockevents (and also for clocksource if both are mixed and John did not
> >> pick it yet) and Thomas pulls from my tree [1].
> >>
> >> If there is no dependency with any other patches of your patchset, which
> &
Commit 54f968d6efdbf7dec36faa44fc11f01b0e4d1990
(tuntap: move socket to tun_file) forgets to set SOCK_ZEROCOPY flag, which will
prevent vhost_net from doing zercopy w/ tap. This patch fixes this by setting
it during file open.
Cc: Michael S. Tsirkin
Signed-off-by: Jason Wang
---
drivers/net
rivers are queued for next release.
>
> I can take it through tip if nobody else wants it :)
Great! I'll queue up 3-6 for the next merge window. That way we avoid
all branch dependencies. :)
thx,
Jason.
--
To unsubscribe from this list: send the line "unsubscribe linux-
+static bool
> +tkt_q_init_contend(int i, arch_spinlock_t *lock, struct __raw_tickets inc)
> +{
> + arch_spinlock_t asold;
> + arch_spinlock_t asnew;
> + struct tkt_q_head *tqhp;
> +
> + /* Initialize the i-th queue header. */
> + tqhp = &tk
ch/arm/plat-orion) and XOR (drivers/dma/mv_xor).
> This also fixes some compiler warnings about now obsolete functions by
> removing those.
>
> Signed-off-by: Sebastian Hesselbarth
> ---
> Cc: Jason Cooper
> Cc: Bartlomiej Zolnierkiewicz
> Cc: Kyungmin
On Mon, Jul 01, 2013 at 05:14:25PM +0200, Sebastian Hesselbarth wrote:
> On 07/01/13 17:04, Jason Cooper wrote:
> >On Mon, Jul 01, 2013 at 04:56:29PM +0200, Sebastian Hesselbarth wrote:
> >>Commit 834cbfe966bece50afded79da8e975d255bf0772
> >> ("drivers/dma:
On Mon, Jul 01, 2013 at 12:42:05PM -0700, Andrew Morton wrote:
> On Mon, 1 Jul 2013 11:04:07 -0400 Jason Cooper wrote:
>
> > On Mon, Jul 01, 2013 at 04:56:29PM +0200, Sebastian Hesselbarth wrote:
> > > Commit 834cbfe966bece50afded79da8e975d255bf0772
> > > ("
ere old code is moved out to and polishes DT board file a little bit.
>
> Signed-off-by: Sebastian Hesselbarth
> ---
> Cc: Russell King
> Cc: Jason Cooper
> Cc: Andrew Lunn
> Cc: Thomas Petazzoni
> Cc: Gregory CLEMENT
> Cc: linux-arm-ker...@lists.infrad
On Tue, Jul 02, 2013 at 02:51:16PM +0200, Sebastian Hesselbarth wrote:
> On 07/02/13 13:53, Jason Cooper wrote:
> >On Tue, Jul 02, 2013 at 01:03:41PM +0200, Sebastian Hesselbarth wrote:
> >>With recent support for true irqchip and clocksource drivers for Orion
> >>SoC
ere old code is moved out to and polishes DT board file a little bit.
>
> Signed-off-by: Sebastian Hesselbarth
> ---
> Changelog:
> v4->v5:
> - do not re-introduce already removed init_dma_coherent_pool_size
> (Reported by Jason Cooper)
>
> @Jason: I did a 'g
On 07/03/2013 06:06 AM, Thomas Zeitlhofer wrote:
> On Tue, Jul 02, 2013 at 06:01:12PM -0300, Fabio Estevam wrote:
>> On Tue, Jul 2, 2013 at 4:59 PM, Thomas Zeitlhofer
>> wrote:
>>> Commit "tuntap: set SOCK_ZEROCOPY flag during open" introduces a
>>> regression which is observed with live migration
Signed-off-by: Jason Wang
---
drivers/net/virtio_net.c |5 +
1 files changed, 1 insertions(+), 4 deletions(-)
diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c
index c9e0038..47b4882 100644
--- a/drivers/net/virtio_net.c
+++ b/drivers/net/virtio_net.c
@@ -901,7 +901,6
>is unneccessary. Remove those board files, Kconfig entries, and
> >>corresponding entries in kirkwood_defconfig.
> >
> >Applying above your set an old quilt cleaning patch Jason asked at some
> >point, it became empty except for arch/arm/mach_kirkwood/common.h in
> >
>is unneccessary. Remove those board files, Kconfig entries, and
> >>corresponding entries in kirkwood_defconfig.
> >
> >Applying above your set an old quilt cleaning patch Jason asked at some
> >point, it became empty except for arch/arm/mach_kirkwood/common.h in
> >
On 07/03/2013 02:11 PM, Thomas Zeitlhofer wrote:
> Hello Jason,
>
> On Wed, Jul 03, 2013 at 10:44:32AM +0800, Jason Wang wrote:
>> On 07/03/2013 06:06 AM, Thomas Zeitlhofer wrote:
>>> On Tue, Jul 02, 2013 at 06:01:12PM -0300, Fabio Estevam wrote:
>>>> On
Orion SoC timer
> >
> > with tip timers/core PR sent, I realized that the above did not make it
> > in. As we have a whole bunch of cleanup and DT conversion patches for
>
> Opps. Fell through the cracks.
No problem.
> > v3.11-rc1 depending on it, would it be ok
On Sun, Jul 07, 2013 at 05:30:31PM +0200, Thomas Gleixner wrote:
> On Sun, 7 Jul 2013, Jason Cooper wrote:
> > Sure, but to be clear, Daniel, please drop this patch from your tree. I
> > have no desire to create an out-of-tree dependency if we can avoid it.
> > It has a ha
On Sun, Jul 07, 2013 at 06:45:15PM +0200, Daniel Lezcano wrote:
> On 07/07/2013 05:30 PM, Thomas Gleixner wrote:
> > On Sun, 7 Jul 2013, Jason Cooper wrote:
> >> Sure, but to be clear, Daniel, please drop this patch from your tree. I
> >> have no desire to create an o
le_dbg(mask, fmt, ...)
> do {
> if (mask & some_module_var)
> debug_something(...)
> } while (0)
>
> It'd be nice to consolidate those in dev_dbg
>
> I'll get 'round to it one day if Jason doesn't.
>
Hi,
I've been a bit
On 06/25/2013 02:03 PM, Joe Perches wrote:
> (Using Jason Baron's most current email address)
>
> On Tue, 2013-06-25 at 10:56 -0700, Joe Perches wrote:
>> On Tue, 2013-06-25 at 10:38 -0700, Greg KH wrote:
>>> On Tue, Jun 25, 2013 at 10:29:50AM -0700, Joe Perches wrote:
static_key_slow_set_true(&control_var);
I see at least 3 users of static_keys in the tree which I think would
benefit from this transformation. The 2 attached with this series, and
the usage in kernel/tracepoint.c.
Thanks,
-Jason
--
To unsubscribe from this list: send the line "
led to this. Each change log should be sufficient to stand
on its own.
Explain why this patch is needed. And it's not about the use of a
simpler API.
It actually fixes a real bug.
Signed-off-by: Jason Baron
---
net/ipv4/udp.c |9 -
net/ipv6/udp.c |9 -
2 files ch
be better to drive the COW break decision off the region's MM
flags? Ie if the memory is mapped read only into the process then you
can keep the COW at the RDMA layer, otherwise you should break
it. That seems more natural than a new flag?
Jason
--
To unsubscribe from this list: send the line
end
unregister
mprotect(WRITABLE)
?
With something like GIFT the app already has to give up writing to the
pages while they are moving, so changing the protection seems in line
with that?
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a
a
> frame number of 0 now. So for each allocation when now_frame == 0 we
> allocate from the dma_pool, not from the free_list. The attached
> patch invalidates the frame number in each iTD before it is sent to
> the scheduler. This fixes the problem without the need to iterate
> over a iTD
Can only be combined with read only access permissions.
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
cessary to ensure that is
impossible.
Jason
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
On Thu, Mar 21, 2013 at 09:15:41PM +0200, Michael S. Tsirkin wrote:
> On Thu, Mar 21, 2013 at 12:41:35PM -0600, Jason Gunthorpe wrote:
> > On Thu, Mar 21, 2013 at 08:16:33PM +0200, Michael S. Tsirkin wrote:
> >
> > > This is the one I find redundant. Since the write w
re why 0xC000 was chosen. Why not 0xD000,
> where the internal registers currently start?
I had the same question earlier but got distracted by other things.
Thanks for bringing it up. Gregory, shouldn't this be 0xD000?
thx,
Jason.
--
To unsubscribe from this list: send the
ive to the ratio of high/low memory, if that gets too
big it struggles, maximizing low memory is desirable - I assume ARM is
basically the same as x86 in this regard?
> So why not map the whole SDRAM above 4GB physical address?
I think the no-MMU boot mode would break if this was done?
Jason
--
On Thu, Mar 21, 2013 at 10:15:23PM +0100, Thomas Petazzoni wrote:
> Dear Jason Gunthorpe,
>
> On Thu, 21 Mar 2013 14:55:45 -0600, Jason Gunthorpe wrote:
>
> > Or, better, locate all the internal registers above 8G and use
> > contiguous DRAM mapping from 0 -> 8GB
>
tions(-)
Whole series applied on top of v3.9-rc3 and tested on dreamplug
(kirkwood DT boot with legacy mv643xx_eth init)
Tested-by: Jason Cooper
also, for the bits changing plat-orion:
Acked-by: Jason Cooper
thx,
Jason.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel&quo
On Fri, Mar 22, 2013 at 03:24:55PM +0100, Florian Fainelli wrote:
> Le 03/22/13 15:14, Jason Cooper a écrit :
> >On Fri, Mar 22, 2013 at 02:39:24PM +0100, Florian Fainelli wrote:
> >>Hi all,
> >>
> >>This patch converts the mv643xx_eth driver to use the m
ed-off-by: David Woodhouse
> Signed-off-by: Jason Cooper
> Signed-off-by: Sebastian Hesselbarth
> ---
> Jason, David,
>
> I tested the patch on Dove and fixed all remaining issues.
>
> Thomas, Gregory, Andrew should test on their platforms, too.
>
> Sebastian
&g
ory */
Reflecting an identity mapping for MMIO - eg MMIO access to 0x60204000
producse a PCI Memory TLP to address 0x60204000 - unless your hardware
is actually doing address translation (then there are other things to
confirm..)
It is usual to have an interrupt-map - have you tested that interrupts
testing, and the rest
will go into the 3.10 merge window.
Cheers,
Jason.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
re, or it is
only called in the KMS (kernel mode setting case). Example:
drivers/tty/vt/vt.c - look at con_debug_enter()
Another option might be to add a variable which allows you to change
the default at compile time, but I would prefer to get the auto detect
code working properly if it is an op
s tcp header to do atr).
So this patches tries to set transport header for packets from untrusted source
(netback, packet, tuntap, macvtap). Plus a fix for better estimation on packet
length for DODGY packet.
Tested on tun/macvtap/packet, compile test on netback.
Jason Wang (5):
macvtap: set tran
skb_flow_dissect() to get l4 offset, if it
fails, just pretend no l4 header.
Cc: Eric Dumazet
Signed-off-by: Jason Wang
---
drivers/net/macvtap.c |9 +
1 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index a449439..acf6450
this patch set the transport header to csum_start for partial checksum
packets, otherwise it first try skb_flow_dissect(), if it fails, just reset the
transport header.
Cc: Eric Dumazet
Signed-off-by: Jason Wang
---
drivers/net/tun.c | 10 ++
1 files changed, 10 insertions(+), 0
) which needs correct l4
header for gso packets.
The patch just reuse the header probed by netback for partial checksum packets
and tries to use skb_flow_dissect() for other cases, if both fail, just pretend
no l4 header.
Cc: Eric Dumazet
Cc: Ian Campbell
Signed-off-by: Jason Wang
---
drivers/net
gso_segs were reset to zero when kernel receive packets from untrusted
source. But we use this zero value to estimate precise packet len which is
wrong. So this patch tries to estimate the correct gso_segs value before using
it in qdisc_pkt_len_init().
Cc: Eric Dumazet
Signed-off-by: Jason Wang
header if skb_flow_dissect() fails.
Cc: Eric Dumazet
Signed-off-by: Jason Wang
---
net/packet/af_packet.c | 21 +
1 files changed, 21 insertions(+), 0 deletions(-)
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index bd0d14c..83fdd0a 100644
--- a/net/packet
> +
> + ret = devm_gpio_request_one(&pdev->dev,
> + port->reset_gpio, GPIOF_DIR_OUT, port->reset_name);
> + if (ret) {
> + if (ret == -EPROBE_DEFER)
> +
rn off debug messages, but that boot parameter is
>> limited to 1023 characters. I'm concerned that if more drivers add
>> dynamic debugging, the distros will eventually run out of space in the
>> dynamic debugging boot parameter. I know Greg was ripping out debugging
>> con
None of its caller use its return value, so let it return void.
Signed-off-by: Jason Wang
---
drivers/vhost/net.c |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 969a859..280ee66 100644
--- a/drivers/vhost/net.c
+++ b
Hi all:
This series tries to unify and simplify vhost codes especially for zerocopy.
Plase review.
Thanks
Jason Wang (6):
vhost_net: make vhost_zerocopy_signal_used() returns void
vhost_net: use vhost_add_used_and_signal_n() in
vhost_zerocopy_signal_used()
vhost: switch to use
Switch to use vhost_add_used_and_signal_n() to avoid multiple calls to
vhost_add_used_and_signal(). With the patch we will call at most 2 times
(consider done_idx warp around) compared to N times w/o this patch.
Signed-off-by: Jason Wang
---
drivers/vhost/net.c | 13 -
1 files
oming from guest. Guest can easily exceeds the limitation.
- We've already check upend_idx != done_idx and switch to non zerocopy then. So
even if all vq->heads were used, we can still does the packet transmission.
So remove this check completely.
Signed-off-by: Jason Wang
---
driver
Let vhost_add_used() to use vhost_add_used_n() to reduce the code duplication.
Signed-off-by: Jason Wang
---
drivers/vhost/vhost.c | 43 ++-
1 files changed, 2 insertions(+), 41 deletions(-)
diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
We used to poll vhost queue before making DMA is done, this is racy if vhost
thread were waked up before marking DMA is done which can result the signal to
be missed. Fix this by always poll the vhost thread before DMA is done.
Signed-off-by: Jason Wang
---
drivers/vhost/net.c |9
Currently, even if the packet length is smaller than VHOST_GOODCOPY_LEN, if
upend_idx != done_idx we still set zcopy_used to true and rollback this choice
later. This could be avoided by determine zerocopy once by checking all
conditions at one time before.
Signed-off-by: Jason Wang
---
drivers
maps physical register addressed onto the same
> virtual addresses.
>
> Signed-off-by: Sebastian Hesselbarth
> ---
> Cc: Russell King
> Cc: Jason Cooper
> Cc: Andrew Lunn
> Cc: Gregory Clement
> Cc: Thomas Petazzoni
> Cc: Arnd Bergmann
> Cc: linux-
bastian Hesselbarth
> ---
> Cc: Rob Herring
> Cc: Pawel Moll
> Cc: Mark Rutland
> Cc: Stephen Warren
> Cc: Ian Campbell
> Cc: Russell King
> Cc: Jason Cooper
> Cc: Andrew Lunn
> Cc: Gregory Clement
> Cc: Thomas Petazzoni
> Cc: Arnd Bergmann
> C
>
> Signed-off-by: Sebastian Hesselbarth
> ---
> Cc: Russell King
> Cc: Jason Cooper
> Cc: Andrew Lunn
> Cc: Gregory Clement
> Cc: Thomas Petazzoni
> Cc: Arnd Bergmann
> Cc: linux-arm-ker...@lists.infradead.org
> Cc: linux-kernel@vger.kernel.org
> ---
> a
bastian Hesselbarth
> ---
> Cc: Rob Herring
> Cc: Pawel Moll
> Cc: Mark Rutland
> Cc: Stephen Warren
> Cc: Ian Campbell
> Cc: Russell King
> Cc: Jason Cooper
> Cc: Andrew Lunn
> Cc: Gregory Clement
> Cc: Thomas Petazzoni
> Cc: Arnd Bergmann
> C
uld you give me a specific example when you build linux-next again?
I'll certainly try to avoid it in the future, but it'll be easier if I
know what 'it' is. ;-)
I'll not change for-next today, and we'll see how it goes.
thx,
Jason.
--
To unsubscribe from this list: se
On Mon, Aug 19, 2013 at 02:09:00PM -0700, Olof Johansson wrote:
> it's this commit:
>
> commit 89602312c5755c87a5ca6ba8ef6b0fce9d510951
> Merge: a0cec78 f23afe2
> Author: Jason Cooper
> AuthorDate: Wed Aug 14 18:55:13 2013 +0000
> Commit: Jason Cooper
> Co
On 08/16/2013 05:54 PM, Michael S. Tsirkin wrote:
> On Fri, Aug 16, 2013 at 01:16:26PM +0800, Jason Wang wrote:
>> > Switch to use vhost_add_used_and_signal_n() to avoid multiple calls to
>> > vhost_add_used_and_signal(). With the patch we will call at most 2 times
>>
On 08/16/2013 05:56 PM, Michael S. Tsirkin wrote:
> On Fri, Aug 16, 2013 at 01:16:27PM +0800, Jason Wang wrote:
>> > Let vhost_add_used() to use vhost_add_used_n() to reduce the code
>> > duplication.
>> >
>> > Signed-off-by: Jason Wang
> Does compiler
On 08/16/2013 06:00 PM, Michael S. Tsirkin wrote:
> On Fri, Aug 16, 2013 at 01:16:29PM +0800, Jason Wang wrote:
>> We used to poll vhost queue before making DMA is done, this is racy if vhost
>> thread were waked up before marking DMA is done which can result the signal
>>
On 08/16/2013 06:02 PM, Michael S. Tsirkin wrote:
> On Fri, Aug 16, 2013 at 01:16:30PM +0800, Jason Wang wrote:
>> We used to limit the max pending DMAs to prevent guest from pinning too many
>> pages. But this could be removed since:
>>
>> - We have the sk_wmem_alloc c
On Sat, Jul 13, 2013 at 12:56:25PM -0700, Olof Johansson wrote:
> On Wed, Jul 10, 2013 at 2:50 PM, Jason Cooper wrote:
> > On Wed, Jul 10, 2013 at 10:08:50PM +0800, Haojian Zhuang wrote:
> >> On Wed, Jul 10, 2013 at 8:24 PM, Jason Cooper wrote:
> >> > On Wed, Ju
if (t == 0) {
> - pr_info("hot_add memory timedout\n");
> - has->ha_end_pfn -= HA_CHUNK;
> - has->covered_end_pfn -= processed_pfn;
> - break;
> - }
> + wait_for_completion_timeout(&a
> -Original Message-
> From: linux-kernel-ow...@vger.kernel.org [mailto:linux-kernel-
> ow...@vger.kernel.org] On Behalf Of Willy Tarreau
> Sent: Wednesday, July 10, 2013 3:27 PM
> To: Rich, Jason
> Cc: linux-kernel@vger.kernel.org
> Subject: Re: Panic at _blk_run_qu
rk.
The main problem with this is that maintainers would need to do the
legwork themselves to figure out how far back a patch is relevant for.
Maybe this would be a feature? ;-)
Unfortunately, since most of the time there's just a single patch, I'm
not sure the cost of the merge
kvm report that has no guest symbols
resolved.
This patch changes the failure to locate kernel modules to be
non-fatal.
Signed-off-by: Jason Wessel
---
tools/perf/util/machine.c |5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/tools/perf/util/machine.c b/tools/perf/util
The hard-coded 8021.q proto will break 802.1ad traffic. So switch to use
vlan->proto.
Cc: Basil Gor
Signed-off-by: Jason Wang
---
drivers/net/macvtap.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 8270b5e..638b
Commit 441ac0fcaadc76ad09771812382345001dd2b813
(macvtap: Convert to using rtnl lock) forget to return what
macvtap_ioctl_set_queue() returns to its caller. This may break multiqueue API
by always falling through to TUNGETFEATURES.
Cc: Vlad Yasevich
Signed-off-by: Jason Wang
---
drivers/net
---
All other % difference results were within a 2% noise range.
Signed-off-by: Jason Low
---
include/linux/sched.h |4
kernel/sched/core.c |3 +++
kernel/sched/fair.c | 26 ++
kernel/sched/sched.h |6 ++
kernel/sysctl.c | 11 ++
501 - 600 of 9518 matches
Mail list logo
