Hi Mauro!
These are the patches which got tested already and
should be good to go. [first batch of patches]
I have another batch with updated patches (dib0700, gp8psk, vp702x)
where I did some more extensive changes to use preallocated memory.
And a small update to the vp7045 patch.
Third batch are the patches to opera1, m920x, dw2102, friio,
a800 which I left as is, for the time beeing.
Regards,
Flo
Florian Mickler (5):
[media] ec168: get rid of on-stack dma buffers
[media] ce6230: get rid of on-stack dma buffer
[media] au6610: get rid of on-stack dma buffer
[media] lmedm04: correct indentation
[media] lmedm04: get rid of on-stack dma buffers
drivers/media/dvb/dvb-usb/au6610.c | 22 ++++++++++++++++------
drivers/media/dvb/dvb-usb/ce6230.c | 11 +++++++++--
drivers/media/dvb/dvb-usb/ec168.c | 18 +++++++++++++++---
drivers/media/dvb/dvb-usb/lmedm04.c | 35 +++++++++++++++++++++++------------
4 files changed, 63 insertions(+), 23 deletions(-)
--
1.7.4.1
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Signed-off-by: Florian Mickler <redacted>
Acked-by: Antti Palosaari <redacted>
Reviewed-by: Antti Palosaari <redacted>
Tested-by: Antti Palosaari <redacted>
---
drivers/media/dvb/dvb-usb/ce6230.c | 11 +++++++++--
1 files changed, 9 insertions(+), 2 deletions(-)
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Signed-off-by: Florian Mickler <redacted>
Acked-by: Antti Palosaari <redacted>
Reviewed-by: Antti Palosaari <redacted>
Tested-by: Antti Palosaari <redacted>
---
drivers/media/dvb/dvb-usb/ec168.c | 18 +++++++++++++++---
1 files changed, 15 insertions(+), 3 deletions(-)
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Tested-By: Malcolm Priestley <tvboxspy@gmail.com>
Signed-off-by: Florian Mickler <redacted>
---
[v2: fix use after free as noted by Malcom]
drivers/media/dvb/dvb-usb/lmedm04.c | 19 +++++++++++++++----
1 files changed, 15 insertions(+), 4 deletions(-)
usb_control_msg initiates (and waits for completion of) a dma transfer using
the supplied buffer. That buffer thus has to be seperately allocated on
the heap.
In lib/dma_debug.c the function check_for_stack even warns about it:
WARNING: at lib/dma-debug.c:866 check_for_stack
Signed-off-by: Florian Mickler <redacted>
Acked-by: Antti Palosaari <redacted>
Reviewed-by: Antti Palosaari <redacted>
Tested-by: Antti Palosaari <redacted>
---
drivers/media/dvb/dvb-usb/au6610.c | 22 ++++++++++++++++------
1 files changed, 16 insertions(+), 6 deletions(-)
@@ -33,8 +33,16 @@ static int au6610_usb_msg(struct dvb_usb_device *d, u8 operation, u8 addr,{intret;u16index;-u8usb_buf[6];/* enough for all known requests,-readreturns5andwrite6bytes*/+u8*usb_buf;++/*+*allocateenoughforallknownrequests,+*readreturns5andwrite6bytes+*/+usb_buf=kmalloc(6,GFP_KERNEL);+if(!usb_buf)+return-ENOMEM;+switch(wlen){case1:index=wbuf[0]<<8;
Hi Mauro!
I just saw that you picked up some patches of mine. What about these?
These are actually tested...
Regards,
Flo
On Sun, 20 Mar 2011 22:50:47 +0100
Florian Mickler [off-list ref] wrote:
Hi Mauro!
These are the patches which got tested already and
should be good to go. [first batch of patches]
I have another batch with updated patches (dib0700, gp8psk, vp702x)
where I did some more extensive changes to use preallocated memory.
And a small update to the vp7045 patch.
Third batch are the patches to opera1, m920x, dw2102, friio,
a800 which I left as is, for the time beeing.
Regards,
Flo
Florian Mickler (5):
[media] ec168: get rid of on-stack dma buffers
[media] ce6230: get rid of on-stack dma buffer
[media] au6610: get rid of on-stack dma buffer
[media] lmedm04: correct indentation
[media] lmedm04: get rid of on-stack dma buffers
drivers/media/dvb/dvb-usb/au6610.c | 22 ++++++++++++++++------
drivers/media/dvb/dvb-usb/ce6230.c | 11 +++++++++--
drivers/media/dvb/dvb-usb/ec168.c | 18 +++++++++++++++---
drivers/media/dvb/dvb-usb/lmedm04.c | 35 +++++++++++++++++++++++------------
4 files changed, 63 insertions(+), 23 deletions(-)
Hi Florian,
Em 30-04-2011 15:54, Florian Mickler escreveu:
Hi Mauro!
I just saw that you picked up some patches of mine. What about these?
These are actually tested...
I'm still in process of applying the pending patches. Due to patchwork.kernel.org
troubles (including the loss of about 270 patches from its SQL database only
recovered yesterday[1]), I have a long backlog. So, I'm gradually applying the remaing
stuff. It will take some time though, and it will depend on patchwork mood, but I intend
to spend some time during this weekend to minimize the backlog.
Cheers,
Mauro
[1] The recover lost the email's body/SOB, so I've wrote a script to use my email
queue to get the data, using patchwork just to mark what patches were already
processed. This increses the time I have to spend on each patch, as I need to run
a script to match the patchwork patch with the patch ID inside my email queue.
Regards,
Flo
On Sun, 20 Mar 2011 22:50:47 +0100
Florian Mickler [off-list ref] wrote:
quoted
Hi Mauro!
These are the patches which got tested already and
should be good to go. [first batch of patches]
I have another batch with updated patches (dib0700, gp8psk, vp702x)
where I did some more extensive changes to use preallocated memory.
And a small update to the vp7045 patch.
Third batch are the patches to opera1, m920x, dw2102, friio,
a800 which I left as is, for the time beeing.
Regards,
Flo
Florian Mickler (5):
[media] ec168: get rid of on-stack dma buffers
[media] ce6230: get rid of on-stack dma buffer
[media] au6610: get rid of on-stack dma buffer
[media] lmedm04: correct indentation
[media] lmedm04: get rid of on-stack dma buffers
drivers/media/dvb/dvb-usb/au6610.c | 22 ++++++++++++++++------
drivers/media/dvb/dvb-usb/ce6230.c | 11 +++++++++--
drivers/media/dvb/dvb-usb/ec168.c | 18 +++++++++++++++---
drivers/media/dvb/dvb-usb/lmedm04.c | 35 +++++++++++++++++++++++------------
4 files changed, 63 insertions(+), 23 deletions(-)
Hi Florian,
Em 30-04-2011 15:54, Florian Mickler escreveu:
quoted
Hi Mauro!
I just saw that you picked up some patches of mine. What about these?
These are actually tested...
I'm still in process of applying the pending patches. Due to patchwork.kernel.org
troubles (including the loss of about 270 patches from its SQL database only
recovered yesterday[1]), I have a long backlog. So, I'm gradually applying the remaing
stuff. It will take some time though, and it will depend on patchwork mood, but I intend
to spend some time during this weekend to minimize the backlog.
Cheers,
Mauro
[1] The recover lost the email's body/SOB, so I've wrote a script to use my email
queue to get the data, using patchwork just to mark what patches were already
processed. This increses the time I have to spend on each patch, as I need to run
a script to match the patchwork patch with the patch ID inside my email queue.
Ah ok, no time pressure over here.. just wanted to make sure
that these don't get lost.
Regards and thanks,
Flo
Hi Florian,
Em 30-04-2011 15:54, Florian Mickler escreveu:
quoted
Hi Mauro!
I just saw that you picked up some patches of mine. What about these?
These are actually tested...
I'm still in process of applying the pending patches. Due to patchwork.kernel.org
troubles (including the loss of about 270 patches from its SQL database only
recovered yesterday[1]), I have a long backlog. So, I'm gradually applying the remaing
stuff. It will take some time though, and it will depend on patchwork mood, but I intend
to spend some time during this weekend to minimize the backlog.
Cheers,
Mauro
[1] The recover lost the email's body/SOB, so I've wrote a script to use my email
queue to get the data, using patchwork just to mark what patches were already
processed. This increses the time I have to spend on each patch, as I need to run
a script to match the patchwork patch with the patch ID inside my email queue.
Ah ok, no time pressure over here.. just wanted to make sure
that these don't get lost.
I think I've applied your series yesterday. Yet, patchwork is currently at bad mood, and it is
not allowing to mark your patches as applied (and the previous series as superseded).
I suspect that I'll need to completely abandon patchwork and work on a new way for handling it
that doesn't depend (nor will update) patchwork.kernel.org.
Mauro.