1/1 
|
Full Buffers
by Unknown on Jun 2, 2005 |
Not available! | ||
|
Good Day
For isochronous Transmission, no Handshaking is made, does this mean that if the ISO buffer is full , packets will drop? is there anyway for controlling the transmission for this issue?
Interrupt Transmission does not have the ability to send NYET Token, in case its buffer is full, what happens? does it retransmit the same packet in the next frame or retries immediately?
Regards
Abukhater
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.opencores.org/pipermail/usb/attachments/20050602/5977f4fe/attachment.html
|
|||
|
Full Buffers
by Unknown on Jun 2, 2005 |
Not available! | ||
|
Hi Abukhater,
True...iso packets are droped and this is called
overrun on fifo.
this can be controller by rate of ISO or using the feedback ISO endpoints.
Interrupt transmission is similar to BULk in all other aspects except rate of
polling is used in interrupt endpoints. So host does retry for the packet.
regards,
pratap
On 6/2/05, M. AbuKhater perocletos at yahoo.com> wrote:
Good Day
For isochronous Transmission, no Handshaking is made, does this mean that if
the ISO buffer is full , packets will drop? is there anyway for controlling
the transmission for this issue?
Interrupt Transmission does not have the ability to send NYET Token, in case
its buffer is full, what happens? does it retransmit the same packet in the
next frame or retries immediately?
Regards
Abukhater
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
-- ********************************************** Pratap Narayan Singh STMicroelectronics Plot No 2&3 Noida ********************************************** |
|||
|
Full Buffers
by Unknown on Jun 2, 2005 |
Not available! | ||
|
Hi Pratap
regarding the interrupt transfer. according to the USB2.0 sepcs P229 Fig 8-38 Interrupt Transmission Does not send NYET Token. so how can we control data transmission during full buffers? retransmission happens only if no ACK token is received.
its a waste of bandwidth to keep retransmitting the Data packet up to 3 times just because the buffer is full.
Regards
Abukhater
Pratap Narayan Singh pratapns at gmail.com> wrote:
Hi Abukhater,
True...iso packets are droped and this is called
overrun on fifo.
this can be controller by rate of ISO or using the feedback ISO endpoints.
Interrupt transmission is similar to BULk in all other aspects except rate of
polling is used in interrupt endpoints. So host does retry for the packet.
regards,
pratap
On 6/2/05, M. AbuKhater
wrote:
Good Day
For isochronous Transmission, no Handshaking is made, does this mean that if
the ISO buffer is full , packets will drop? is there anyway for controlling
the transmission for this issue?
Interrupt Transmission does not have the ability to send NYET Token, in case
its buffer is full, what happens? does it retransmit the same packet in the
next frame or retries immediately?
Regards
Abukhater
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
--
**********************************************
Pratap Narayan Singh
STMicroelectronics
Plot No 2&3
Noida
**********************************************
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.opencores.org/pipermail/usb/attachments/20050602/5df4141a/attachment.html
|
|||
|
Full Buffers
by Unknown on Jun 2, 2005 |
Not available! | ||
|
Dear Pratap,
Dear AbuKhater, I see that both of you seem to work with some USB2.0 IP cores. Does any of you have any really world experience with the cores?? I am considering to use an USB2.0 core for a future design. Well, I am only in the initial planning phase. All I need to have is a BULK transfer, but the transfers have to be pretty fast. I will have to transfer VGA images at 30 frames per second with each pixel having 16bit. Now, is there somewhere something that allows me to verify the cores? What are your experience with it? Best regards Georg On Thu, 2005-06-02 at 03:01 -0700, M. AbuKhater wrote:
Hi Pratap
regarding the interrupt transfer. according to the USB2.0 sepcs P229
Fig 8-38 Interrupt Transmission Does not send NYET Token. so how can
we control data transmission during full buffers? retransmission
happens only if no ACK token is received.
its a waste of bandwidth to keep retransmitting the Data packet up to
3 times just because the buffer is full.
Regards
Abukhater
Pratap Narayan Singh pratapns at gmail.com> wrote:
Hi Abukhater,
True...iso packets are droped and this is called
overrun on fifo.
this can be controller by rate of ISO or using the feedback
ISO endpoints.
Interrupt transmission is similar to BULk in all other aspects
except rate of
polling is used in interrupt endpoints. So host does retry for
the packet.
regards,
pratap
On 6/2/05, M. AbuKhater wrote:
> Good Day
> For isochronous Transmission, no Handshaking is made, does this mean that if
> the ISO buffer is full , packets will drop? is there anyway
for controlling
> the transmission for this issue?
> > Interrupt Transmission does not have the ability to send NYET Token, in case
> its buffer is full, what happens? does it retransmit the
same packet in the
> next frame or retries immediately?
> > Regards > Abukhater > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
>
>
--
**********************************************
Pratap Narayan Singh
STMicroelectronics
Plot No 2&3
Noida
**********************************************
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
|
|||
|
Full Buffers
by Unknown on Jun 2, 2005 |
Not available! | ||
|
I am currently using the OpenCore USB2.0 core.
The implementation is good and is easy to work with. However, there
are a few issues that I had to fix in order to get it to work
properly. Namely, changes were required to the PE for ACK/NACKing and
zero-length packet handling, as well as the STALLing mechanism.
I currently have it working on a custom Linux platform (FPGA based)
as an ethernet gadget and as a file backed storage gadget so I've put
it through it's paces a good bit. My confidence level in the core in
general is high.
I'm hoping that I will be able to release these changes back to
OpenCores in short order.
You should be able to get VGA (640 x 480) @ 16-bit @ 30fps across a
USB2.0 high-speed link without much trouble.
640 * 480 * 2 * 30 = 18.432 MBps which should be easily achievable
with USB2.0 running in high-speed mode.
On 6/2/05, CMOS Vision admin at cmosvision.com> wrote:
Dear Pratap,
Dear AbuKhater, I see that both of you seem to work with some USB2.0 IP cores. Does any of you have any really world experience with the cores?? I am considering to use an USB2.0 core for a future design. Well, I am only in the initial planning phase. All I need to have is a BULK transfer, but the transfers have to be pretty fast. I will have to transfer VGA images at 30 frames per second with each pixel having 16bit. Now, is there somewhere something that allows me to verify the cores? What are your experience with it? Best regards Georg On Thu, 2005-06-02 at 03:01 -0700, M. AbuKhater wrote:
> Hi Pratap
> regarding the interrupt transfer. according to the USB2.0 sepcs P229
> Fig 8-38 Interrupt Transmission Does not send NYET Token. so how can
> we control data transmission during full buffers? retransmission
> happens only if no ACK token is received.
> its a waste of bandwidth to keep retransmitting the Data packet up to
> 3 times just because the buffer is full.
>
> Regards
> Abukhater
>
>
> Pratap Narayan Singh pratapns at gmail.com> wrote:
> Hi Abukhater,
> True...iso packets are droped and this is called
> overrun on fifo.
> this can be controller by rate of ISO or using the feedback
> ISO endpoints.
> Interrupt transmission is similar to BULk in all other aspects
> except rate of
> polling is used in interrupt endpoints. So host does retry for
> the packet.
> regards,
> pratap
>
> On 6/2/05, M. AbuKhater wrote:
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
> Good Day
> For isochronous Transmission, no Handshaking is made, does > this mean that if
> the ISO buffer is full , packets will drop? is there anyway
> for controlling
> the transmission for this issue?
> > Interrupt Transmission does not have the ability to send > NYET Token, in case
> its buffer is full, what happens? does it retransmit the
> same packet in the
> next frame or retries immediately?
> > Regards > Abukhater > > __________________________________________________ > Do You Yahoo!? > Tired of spam? Yahoo! Mail has the best spam protection > around
> http://mail.yahoo.com
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
>
>
>
>
> --
> **********************************************
> Pratap Narayan Singh
> STMicroelectronics
> Plot No 2&3
> Noida
> **********************************************
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
|
|||
|
Full Buffers
by Unknown on Jun 2, 2005 |
Not available! | ||
|
Dear Jeremy,
thanks for this kind replay. It is great to read this. BTW, what type of physical interface chip do you use with it?? Best regards Georg On Thu, 2005-06-02 at 10:08 -0600, Jeremy Fillingim wrote:
I am currently using the OpenCore USB2.0 core.
The implementation is good and is easy to work with. However, there
are a few issues that I had to fix in order to get it to work
properly. Namely, changes were required to the PE for ACK/NACKing and
zero-length packet handling, as well as the STALLing mechanism.
I currently have it working on a custom Linux platform (FPGA based)
as an ethernet gadget and as a file backed storage gadget so I've put
it through it's paces a good bit. My confidence level in the core in
general is high.
I'm hoping that I will be able to release these changes back to
OpenCores in short order.
You should be able to get VGA (640 x 480) @ 16-bit @ 30fps across a
USB2.0 high-speed link without much trouble.
640 * 480 * 2 * 30 = 18.432 MBps which should be easily achievable
with USB2.0 running in high-speed mode.
On 6/2/05, CMOS Vision admin at cmosvision.com> wrote:
> Dear Pratap,
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
> Dear AbuKhater, > > I see that both of you seem to work with some USB2.0 IP cores. > Does any of you have any really world experience with the cores?? > > I am considering to use an USB2.0 core for a future design. > Well, I am only in the initial planning phase. > All I need to have is a BULK transfer, but the transfers have to be > pretty fast. I will have to transfer VGA images at 30 frames per second > with each pixel having 16bit. > Now, is there somewhere something that allows me to verify the cores? > > What are your experience with it? > > Best regards > > Georg > > > On Thu, 2005-06-02 at 03:01 -0700, M. AbuKhater wrote:
> Hi Pratap
> regarding the interrupt transfer. according to the USB2.0 sepcs P229
> Fig 8-38 Interrupt Transmission Does not send NYET Token. so how can
> we control data transmission during full buffers? retransmission
> happens only if no ACK token is received.
> its a waste of bandwidth to keep retransmitting the Data packet up to
> 3 times just because the buffer is full.
>
> Regards
> Abukhater
>
>
> Pratap Narayan Singh pratapns at gmail.com> wrote:
> Hi Abukhater,
> True...iso packets are droped and this is called
> overrun on fifo.
> this can be controller by rate of ISO or using the feedback
> ISO endpoints.
> Interrupt transmission is similar to BULk in all other aspects
> except rate of
> polling is used in interrupt endpoints. So host does retry for
> the packet.
> regards,
> pratap
>
> On 6/2/05, M. AbuKhater wrote:
>
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
>
> > Good Day
> > For isochronous Transmission, no Handshaking is made, does > this mean that if
> > the ISO buffer is full , packets will drop? is there anyway
> for controlling
> > the transmission for this issue?
> > > > Interrupt Transmission does not have the ability to send > NYET Token, in case
> > its buffer is full, what happens? does it retransmit the
> same packet in the
> > next frame or retries immediately?
> > > > Regards > > Abukhater > > > > __________________________________________________ > > Do You Yahoo!? > > Tired of spam? Yahoo! Mail has the best spam protection > around
> > http://mail.yahoo.com
> > _______________________________________________
> > http://www.opencores.org/mailman/listinfo/usb
> >
> >
>
>
> --
> **********************************************
> Pratap Narayan Singh
> STMicroelectronics
> Plot No 2&3
> Noida
> **********************************************
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
|
|||
|
Full Buffers
by Unknown on Jun 3, 2005 |
Not available! | ||
|
Hi Jeremy,
Did you ever tried to verify it with some CRV environment ?
pratap
On 6/2/05, Jeremy Fillingim jfilling at gmail.com> wrote:
I am currently using the OpenCore USB2.0 core.
The implementation is good and is easy to work with. However, there
are a few issues that I had to fix in order to get it to work
properly. Namely, changes were required to the PE for ACK/NACKing and
zero-length packet handling, as well as the STALLing mechanism.
I currently have it working on a custom Linux platform (FPGA based)
as an ethernet gadget and as a file backed storage gadget so I've put
it through it's paces a good bit. My confidence level in the core in
general is high.
I'm hoping that I will be able to release these changes back to
OpenCores in short order.
You should be able to get VGA (640 x 480) @ 16-bit @ 30fps across a
USB2.0 high-speed link without much trouble.
640 * 480 * 2 * 30 = 18.432 MBps which should be easily achievable
with USB2.0 running in high-speed mode.
On 6/2/05, CMOS Vision admin at cmosvision.com> wrote:
> Dear Pratap,
_______________________________________________
http://www.opencores.org/mailman/listinfo/usb
> Dear AbuKhater, > > I see that both of you seem to work with some USB2.0 IP cores. > Does any of you have any really world experience with the cores?? > > I am considering to use an USB2.0 core for a future design. > Well, I am only in the initial planning phase. > All I need to have is a BULK transfer, but the transfers have to be > pretty fast. I will have to transfer VGA images at 30 frames per second > with each pixel having 16bit. > Now, is there somewhere something that allows me to verify the cores? > > What are your experience with it? > > Best regards > > Georg > > > On Thu, 2005-06-02 at 03:01 -0700, M. AbuKhater wrote:
> Hi Pratap
> regarding the interrupt transfer. according to the USB2.0 sepcs P229
> Fig 8-38 Interrupt Transmission Does not send NYET Token. so how can
> we control data transmission during full buffers? retransmission
> happens only if no ACK token is received.
> its a waste of bandwidth to keep retransmitting the Data packet up to
> 3 times just because the buffer is full.
>
> Regards
> Abukhater
>
>
> Pratap Narayan Singh pratapns at gmail.com> wrote:
> Hi Abukhater,
> True...iso packets are droped and this is called
> overrun on fifo.
> this can be controller by rate of ISO or using the feedback
> ISO endpoints.
> Interrupt transmission is similar to BULk in all other aspects
> except rate of
> polling is used in interrupt endpoints. So host does retry for
> the packet.
> regards,
> pratap
>
> On 6/2/05, M. AbuKhater wrote:
>
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
>
> > Good Day
> > For isochronous Transmission, no Handshaking is made, does > this mean that if
> > the ISO buffer is full , packets will drop? is there anyway
> for controlling
> > the transmission for this issue?
> > > > Interrupt Transmission does not have the ability to send > NYET Token, in case
> > its buffer is full, what happens? does it retransmit the
> same packet in the
> > next frame or retries immediately?
> > > > Regards > > Abukhater > > > > __________________________________________________ > > Do You Yahoo!? > > Tired of spam? Yahoo! Mail has the best spam protection > around
> > http://mail.yahoo.com
> > _______________________________________________
> > http://www.opencores.org/mailman/listinfo/usb
> >
> >
>
>
> --
> **********************************************
> Pratap Narayan Singh
> STMicroelectronics
> Plot No 2&3
> Noida
> **********************************************
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam? Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
> _______________________________________________
> http://www.opencores.org/mailman/listinfo/usb
-- ********************************************** Pratap Narayan Singh STMicroelectronics Plot No 2&3 Noida ********************************************** |
|||
1/1