Skip to contentSkip to conversation

playback write error (DMA or IRQ trouble)

DISCUSSION Harsha priya gupta 2 participants 12 messages Jun 2, 2008

lore
Share & tools
lore
#0 lore raw

Can anyone give me a clue as to when i would get such an error?

Signature · 1 line
-Harsha
#1 lore raw

At Mon, 2 Jun 2008 12:39:31 +0530,

Quoted 2 lines · Harsha priya gupta

Can anyone give me a clue as to when i would get such an error?

... only if you give more clue what exactly you did.

In general, it implies that an interrupt isn't issued properly at PCM period boundary.

Takashi

#2 lore raw

I implemented the copy function and immediately transfered the user block data to the hardware.

Correct me if am wrong;
.pointer implementation - passes the current buffer pointer. When the .pointer function returns the size of the buffer = user buffer size logically I need to expect the hardware to send an interrupt because buffer is consumed and I should call snd_pcm_period_elapsed after that.

what would happen if i call the snd_pcm_period_elapsed from the pointer function once the buffer is consumed from hardware. Would that be right? This is what i am trying to do

Quoted 13 lines · Takashi Iwai

At Mon, 2 Jun 2008 12:39:31 +0530,

Quoted 2 lines · Harsha priya gupta

Can anyone give me a clue as to when i would get such an error?

... only if you give more clue what exactly you did.

In general, it implies that an interrupt isn't issued properly at PCM period boundary.

Takashi

Signature · 1 line
-Harsha
#3 lore raw

At Mon, 2 Jun 2008 13:26:01 +0530,

Quoted 10 lines · Harsha priya gupta

I implemented the copy function and immediately transfered the user block data to the hardware.

Correct me if am wrong;
.pointer implementation - passes the current buffer pointer. When the .pointer function returns the size of the buffer = user buffer size logically I need to expect the hardware to send an interrupt because buffer is consumed and I should call snd_pcm_period_elapsed after that.

Quoted 3 lines

what would happen if i call the snd_pcm_period_elapsed from the pointer function once the buffer is consumed from hardware. Would that be right? This is what i am trying to do

The logic is reversed.
The pointer callback is a passive one that does nothing but returning the current h/w buffer position. This is called either from snd_pcm_period_elapsed() or at the PCM status update.

You must call snd_pcm_period_elapsed() somewhere in your driver *explicitly* at the timing that one period is finished. Usually, this is done in an IRQ handler the h/w generates at the period ("fragment", "half-buffer", or whatever) boundary.

And note that the valid value from the pointer callback is between 0 and buffer_size-1 as it handles the buffer as a ring-buffer. The value buffer_size is invalid.

Takashi

Quoted 18 lines · Takashi Iwai

At Mon, 2 Jun 2008 12:39:31 +0530,
Harsha priya gupta wrote:
>
> Can anyone give me a clue as to when i would get such an error?

... only if you give more clue what exactly you did.

In general, it implies that an interrupt isn't issued properly at PCM period boundary.

Takashi

--
-Harsha

#4 lore raw

Quick question

Quoted 1 line

From my copy function after I pass the buffer to HW, what would happen if i

call snd_pcm_period_elapsed.

Quoted 56 lines · Takashi Iwai

At Mon, 2 Jun 2008 13:26:01 +0530,

Quoted 2 lines · Harsha priya gupta

I implemented the copy function and immediately transfered the user block

data

Quoted 4 lines

to the hardware.

Correct me if am wrong;
.pointer implementation - passes the current buffer pointer. When the

.pointer

Quoted 1 line

function returns the size of the buffer = user buffer size logically I

need to

Quoted 5 lines

expect the hardware to send an interrupt because buffer is consumed and I should call snd_pcm_period_elapsed after that.

what would happen if i call the snd_pcm_period_elapsed from the pointer function once the buffer is consumed from hardware. Would that be right?

This

Quoted 1 line

is what i am trying to do

The logic is reversed.
The pointer callback is a passive one that does nothing but returning the current h/w buffer position. This is called either from snd_pcm_period_elapsed() or at the PCM status update.

You must call snd_pcm_period_elapsed() somewhere in your driver *explicitly* at the timing that one period is finished. Usually, this is done in an IRQ handler the h/w generates at the period ("fragment", "half-buffer", or whatever) boundary.

And note that the valid value from the pointer callback is between 0 and buffer_size-1 as it handles the buffer as a ring-buffer. The value buffer_size is invalid.

Takashi

Quoted 18 lines · Takashi Iwai

At Mon, 2 Jun 2008 12:39:31 +0530,
Harsha priya gupta wrote:
>
> Can anyone give me a clue as to when i would get such an error?

... only if you give more clue what exactly you did.

In general, it implies that an interrupt isn't issued properly at PCM period boundary.

Takashi

--
-Harsha

Signature · 1 line
-Harsha
#5 lore raw

At Mon, 2 Jun 2008 14:33:14 +0530,

Quoted 5 lines · Harsha priya gupta

Quick question

From my copy function after I pass the buffer to HW, what would happen if i call snd_pcm_period_elapsed.

It's invalid and a misdesign.

I guess you are misunderstanding about when to callsnd_pcm_period_elapsed(). snd_pcm_period_elapsed() is called when one period of samples on the hardware is *processed*. It doesn't mean that the samples are transferred to the hardware.

Suppose that you have period_size = 48000 (frames) for 48kHz samples. Then, the first snd_pcm_period_epased() shall be called just one second after starting the PCM stream. The second call be another one second later, and so on. It doesn't matter how quick the copy to h/w is done (via copy callback).

Takashi

Quoted 62 lines · Takashi Iwai

At Mon, 2 Jun 2008 13:26:01 +0530,
Harsha priya gupta wrote:
>
> I implemented the copy function and immediately transfered the user block data
> to the hardware.
>
> Correct me if am wrong;
> .pointer implementation - passes the current buffer pointer. When the .pointer
> function returns the size of the buffer = user buffer size logically I need to
> expect the hardware to send an interrupt because buffer is consumed and I > should call snd_pcm_period_elapsed after that.
>
> what would happen if i call the snd_pcm_period_elapsed from the pointer > function once the buffer is consumed from hardware. Would that be right?
This
> is what i am trying to do

The logic is reversed.
The pointer callback is a passive one that does nothing but returning the current h/w buffer position. This is called either from snd_pcm_period_elapsed() or at the PCM status update.

You must call snd_pcm_period_elapsed() somewhere in your driver *explicitly* at the timing that one period is finished. Usually, this is done in an IRQ handler the h/w generates at the period ("fragment", "half-buffer", or whatever) boundary.

And note that the valid value from the pointer callback is between 0 and buffer_size-1 as it handles the buffer as a ring-buffer. The value buffer_size is invalid.

Takashi

> On Mon, Jun 2, 2008 at 1:02 PM, Takashi Iwai <[email protected]> wrote:
>
> At Mon, 2 Jun 2008 12:39:31 +0530,
> Harsha priya gupta wrote:
> >
> > Can anyone give me a clue as to when i would get such an error?
>
> ... only if you give more clue what exactly you did.
>
> In general, it implies that an interrupt isn't issued properly at PCM
> period boundary.
>
> Takashi
>
> --
> -Harsha
>
>

--
-Harsha

#6 lore raw

Say if my hardware is such that it shall interrupt only after it has processed entire sample and not ever period or sample. What will ensure that i get my next buffer down? Will calling the snd_pcm_period_elapsed in the interrupt function help?

Quoted 99 lines · Takashi Iwai

At Mon, 2 Jun 2008 14:33:14 +0530,

Quoted 4 lines · Harsha priya gupta

Quick question

From my copy function after I pass the buffer to HW, what would happen if

i

Quoted 1 line

call snd_pcm_period_elapsed.

It's invalid and a misdesign.

I guess you are misunderstanding about when to callsnd_pcm_period_elapsed(). snd_pcm_period_elapsed() is called when one period of samples on the hardware is *processed*. It doesn't mean that the samples are transferred to the hardware.

Suppose that you have period_size = 48000 (frames) for 48kHz samples. Then, the first snd_pcm_period_epased() shall be called just one second after starting the PCM stream. The second call be another one second later, and so on. It doesn't matter how quick the copy to h/w is done (via copy callback).

Takashi

Quoted 7 lines · Takashi Iwai

At Mon, 2 Jun 2008 13:26:01 +0530,
Harsha priya gupta wrote:
>
> I implemented the copy function and immediately transfered the user

block

Quoted 5 lines

data
> to the hardware.
>
> Correct me if am wrong;
> .pointer implementation - passes the current buffer pointer. When

the

Quoted 2 lines

.pointer
> function returns the size of the buffer = user buffer size

logically I

Quoted 2 lines

need to
> expect the hardware to send an interrupt because buffer is consumed

and I

Quoted 3 lines
Quoted 3 lines

should call snd_pcm_period_elapsed after that.

what would happen if i call the snd_pcm_period_elapsed from the

pointer

Quoted 1 line
Quoted 1 line

function once the buffer is consumed from hardware. Would that be

right?

Quoted 10 lines

This
> is what i am trying to do

The logic is reversed.
The pointer callback is a passive one that does nothing but returning the current h/w buffer position. This is called either from snd_pcm_period_elapsed() or at the PCM status update.

You must call snd_pcm_period_elapsed() somewhere in your driver *explicitly* at the timing that one period is finished. Usually,

this

Quoted 1 line

is done in an IRQ handler the h/w generates at the period

("fragment",

Quoted 9 lines

"half-buffer", or whatever) boundary.

And note that the valid value from the pointer callback is between 0 and buffer_size-1 as it handles the buffer as a ring-buffer. The value buffer_size is invalid.

Takashi

> On Mon, Jun 2, 2008 at 1:02 PM, Takashi Iwai <[email protected]>

Quoted 5 lines · wrote:
Quoted 5 lines

At Mon, 2 Jun 2008 12:39:31 +0530,
Harsha priya gupta wrote:
>
> Can anyone give me a clue as to when i would get such an

error?

Quoted 4 lines
Quoted 4 lines

... only if you give more clue what exactly you did.

In general, it implies that an interrupt isn't issued properly

at PCM

Quoted 13 lines
Quoted 8 lines

period boundary.

Takashi

--
-Harsha

--
-Harsha

Signature · 1 line
-Harsha
#7 lore raw

At Mon, 2 Jun 2008 15:06:03 +0530,

Quoted 5 lines · Harsha priya gupta

Say if my hardware is such that it shall interrupt only after it has processed entire sample and not ever period or sample. What will ensure that i get my next buffer down? Will calling the snd_pcm_period_elapsed in the interrupt function help?

So, your hardware has only a single ring buffer and can issue an interrupt only at the end of the buffer?

If so, you might need to seek for another interrupt source, such as a system timer.

Takashi

Quoted 103 lines · Takashi Iwai

At Mon, 2 Jun 2008 14:33:14 +0530,
Harsha priya gupta wrote:
>
> Quick question
>
> From my copy function after I pass the buffer to HW, what would happen if i
> call snd_pcm_period_elapsed.

It's invalid and a misdesign.

I guess you are misunderstanding about when to callsnd_pcm_period_elapsed(). snd_pcm_period_elapsed() is called when one period of samples on the hardware is *processed*. It doesn't mean that the samples are transferred to the hardware.

Suppose that you have period_size = 48000 (frames) for 48kHz samples. Then, the first snd_pcm_period_epased() shall be called just one second after starting the PCM stream. The second call be another one second later, and so on. It doesn't matter how quick the copy to h/w is done (via copy callback).

Takashi

>
> On Mon, Jun 2, 2008 at 1:37 PM, Takashi Iwai <[email protected]> wrote:
>
> At Mon, 2 Jun 2008 13:26:01 +0530,
> Harsha priya gupta wrote:
> >
> > I implemented the copy function and immediately transfered the user
block
> data
> > to the hardware.
> >
> > Correct me if am wrong;
> > .pointer implementation - passes the current buffer pointer. When
the
> .pointer
> > function returns the size of the buffer = user buffer size
logically I
> need to
> > expect the hardware to send an interrupt because buffer is consumed
and I
> > should call snd_pcm_period_elapsed after that.
> >
> > what would happen if i call the snd_pcm_period_elapsed from the
pointer
> > function once the buffer is consumed from hardware. Would that be
right?
> This
> > is what i am trying to do
>
> The logic is reversed.
> The pointer callback is a passive one that does nothing but returning
> the current h/w buffer position. This is called either from
> snd_pcm_period_elapsed() or at the PCM status update.
>
> You must call snd_pcm_period_elapsed() somewhere in your driver
> *explicitly* at the timing that one period is finished. Usually,
this
> is done in an IRQ handler the h/w generates at the period
("fragment",
> "half-buffer", or whatever) boundary.
>
> And note that the valid value from the pointer callback is between 0
> and buffer_size-1 as it handles the buffer as a ring-buffer. The
> value buffer_size is invalid.
>
> Takashi
>
> > On Mon, Jun 2, 2008 at 1:02 PM, Takashi Iwai <[email protected]> wrote:
> >
> > At Mon, 2 Jun 2008 12:39:31 +0530,
> > Harsha priya gupta wrote:
> > >
> > > Can anyone give me a clue as to when i would get such an
error?
> >
> > ... only if you give more clue what exactly you did.
> >
> > In general, it implies that an interrupt isn't issued properly
at PCM
> > period boundary.
> >
> > Takashi
> >
> > --
> > -Harsha
> >
> >
>
> --
> -Harsha
>
>

--
-Harsha

#8 lore raw

Is there any way to avoid the timer part? Because i do not want the cpu to bother with anything till the hardware has done its part. after the hardware interrupts, can i manipulate someway to get the next buffer rather than interrupting every 1 sec to say that the period has been processed.

Quoted 140 lines · Takashi Iwai

At Mon, 2 Jun 2008 15:06:03 +0530,

Quoted 2 lines · Harsha priya gupta

Say if my hardware is such that it shall interrupt only after it has

processed

Quoted 1 line

entire sample and not ever period or sample. What will ensure that i get

my

Quoted 1 line

next buffer down? Will calling the snd_pcm_period_elapsed in the

interrupt

Quoted 1 line

function help?

So, your hardware has only a single ring buffer and can issue an interrupt only at the end of the buffer?

If so, you might need to seek for another interrupt source, such as a system timer.

Takashi

Quoted 8 lines · Takashi Iwai

At Mon, 2 Jun 2008 14:33:14 +0530,
Harsha priya gupta wrote:
>
> Quick question
>
> From my copy function after I pass the buffer to HW, what would

happen if

Quoted 7 lines

i
> call snd_pcm_period_elapsed.

It's invalid and a misdesign.

I guess you are misunderstanding about when to callsnd_pcm_period_elapsed(). snd_pcm_period_elapsed() is called

when

Quoted 1 line

one period of samples on the hardware is *processed*. It doesn't

mean

Quoted 12 lines

that the samples are transferred to the hardware.

Suppose that you have period_size = 48000 (frames) for 48kHz samples. Then, the first snd_pcm_period_epased() shall be called just one second after starting the PCM stream. The second call be another one second later, and so on. It doesn't matter how quick the copy to h/w is done (via copy callback).

Takashi

>
> On Mon, Jun 2, 2008 at 1:37 PM, Takashi Iwai <[email protected]>

Quoted 5 lines · wrote:
Quoted 5 lines

At Mon, 2 Jun 2008 13:26:01 +0530,
Harsha priya gupta wrote:
>
> I implemented the copy function and immediately transfered

the user

Quoted 6 lines

block
> data
> > to the hardware.
> >
> > Correct me if am wrong;
> > .pointer implementation - passes the current buffer pointer.

When

Quoted 6 lines

the
> .pointer
> > function returns the size of the buffer = user buffer size
logically I
> need to
> > expect the hardware to send an interrupt because buffer is

consumed

Quoted 4 lines

and I
> > should call snd_pcm_period_elapsed after that.
> >
> > what would happen if i call the snd_pcm_period_elapsed from

the

Quoted 2 lines

pointer
> > function once the buffer is consumed from hardware. Would

that be

Quoted 6 lines

right?
> This
> > is what i am trying to do
>
> The logic is reversed.
> The pointer callback is a passive one that does nothing but

returning

Quoted 5 lines
Quoted 5 lines

the current h/w buffer position. This is called either from snd_pcm_period_elapsed() or at the PCM status update.

You must call snd_pcm_period_elapsed() somewhere in your driver *explicitly* at the timing that one period is finished.

Usually,

Quoted 6 lines

this
> is done in an IRQ handler the h/w generates at the period
("fragment",
> "half-buffer", or whatever) boundary.
>
> And note that the valid value from the pointer callback is

between 0

Quoted 1 line
Quoted 1 line

and buffer_size-1 as it handles the buffer as a ring-buffer.

The

Quoted 5 lines
Quoted 5 lines

value buffer_size is invalid.

Takashi

> On Mon, Jun 2, 2008 at 1:02 PM, Takashi Iwai <[email protected]>

Quoted 5 lines · wrote:
Quoted 5 lines
Quoted 5 lines

At Mon, 2 Jun 2008 12:39:31 +0530,
Harsha priya gupta wrote:
>
> Can anyone give me a clue as to when i would get such

an

Quoted 5 lines

error?
> >
> > ... only if you give more clue what exactly you did.
> >
> > In general, it implies that an interrupt isn't issued

properly

Quoted 19 lines
    at PCM
    >     >     period boundary.
    >     >
    >     >     Takashi
    >     >
    >     > --
    >     > -Harsha
    >     >
    >     >
    >
    > --
    > -Harsha
    >
    >

--
-Harsha

Signature · 1 line
-Harsha
#9 lore raw

At Mon, 2 Jun 2008 15:16:06 +0530,

Quoted 3 lines · Harsha priya gupta

Is there any way to avoid the timer part? Because i do not want the cpu to bother with anything till the hardware has done its part.

But your hardware doesn't do what we need. A basic part is missing.

Quoted 3 lines

after the hardware
interrupts, can i manipulate someway to get the next buffer rather than interrupting every 1 sec to say that the period has been processed.

Well, how would you handle "the next buffer"?

Suppose all samples on the buffer are processed and an irq is issued. At this moment, you have no data on the buffer. It means that the device is already in a buffer-underrun error state.

The period-based transfer is the basis of ALSA PCM transfer model. In this way, you can write programs requiring a precise timing control, too.

Takashi

Quoted 145 lines · Takashi Iwai

At Mon, 2 Jun 2008 15:06:03 +0530,
Harsha priya gupta wrote:
>
> Say if my hardware is such that it shall interrupt only after it has processed
> entire sample and not ever period or sample. What will ensure that i get my
> next buffer down? Will calling the snd_pcm_period_elapsed in the interrupt
> function help?

So, your hardware has only a single ring buffer and can issue an interrupt only at the end of the buffer?

If so, you might need to seek for another interrupt source, such as a system timer.

Takashi

> On Mon, Jun 2, 2008 at 2:54 PM, Takashi Iwai <[email protected]> wrote:
>
> At Mon, 2 Jun 2008 14:33:14 +0530,
> Harsha priya gupta wrote:
> >
> > Quick question
> >
> > From my copy function after I pass the buffer to HW, what would
happen if
> i
> > call snd_pcm_period_elapsed.
>
> It's invalid and a misdesign.
>
> I guess you are misunderstanding about when to
> callsnd_pcm_period_elapsed(). snd_pcm_period_elapsed() is called
when
> one period of samples on the hardware is *processed*. It doesn't
mean
> that the samples are transferred to the hardware.
>
> Suppose that you have period_size = 48000 (frames) for 48kHz samples.
> Then, the first snd_pcm_period_epased() shall be called just one
> second after starting the PCM stream. The second call be another one
> second later, and so on. It doesn't matter how quick the copy to h/w
> is done (via copy callback).
>
> Takashi
>
> >
> > On Mon, Jun 2, 2008 at 1:37 PM, Takashi Iwai <[email protected]> wrote:
> >
> > At Mon, 2 Jun 2008 13:26:01 +0530,
> > Harsha priya gupta wrote:
> > >
> > > I implemented the copy function and immediately transfered
the user
> block
> > data
> > > to the hardware.
> > >
> > > Correct me if am wrong;
> > > .pointer implementation - passes the current buffer pointer.
When
> the
> > .pointer
> > > function returns the size of the buffer = user buffer size
> logically I
> > need to
> > > expect the hardware to send an interrupt because buffer is
consumed
> and I
> > > should call snd_pcm_period_elapsed after that.
> > >
> > > what would happen if i call the snd_pcm_period_elapsed from
the
> pointer
> > > function once the buffer is consumed from hardware. Would
that be
> right?
> > This
> > > is what i am trying to do
> >
> > The logic is reversed.
> > The pointer callback is a passive one that does nothing but
returning
> > the current h/w buffer position. This is called either from
> > snd_pcm_period_elapsed() or at the PCM status update.
> >
> > You must call snd_pcm_period_elapsed() somewhere in your driver
> > *explicitly* at the timing that one period is finished.
Usually,
> this
> > is done in an IRQ handler the h/w generates at the period
> ("fragment",
> > "half-buffer", or whatever) boundary.
> >
> > And note that the valid value from the pointer callback is
between 0
> > and buffer_size-1 as it handles the buffer as a ring-buffer.
The
> > value buffer_size is invalid.
> >
> > Takashi
> >
> > > On Mon, Jun 2, 2008 at 1:02 PM, Takashi Iwai <[email protected]>
wrote:
> > >
> > > At Mon, 2 Jun 2008 12:39:31 +0530,
> > > Harsha priya gupta wrote:
> > > >
> > > > Can anyone give me a clue as to when i would get such
an
> error?
> > >
> > > ... only if you give more clue what exactly you did.
> > >
> > > In general, it implies that an interrupt isn't issued
properly
> at PCM
> > > period boundary.
> > >
> > > Takashi
> > >
> > > --
> > > -Harsha
> > >
> > >
> >
> > --
> > -Harsha
> >
> >
>
> --
> -Harsha
>
>

--
-Harsha

#10 lore raw

OK. So can i assume that its compulsory to raise at least a timer interrupt every 1 second (for 48K sample/ 48 kHz). if its 96Khz, then i would have to raise the timer interrupt every 2 seconds... correct?

Quoted 198 lines · Takashi Iwai

At Mon, 2 Jun 2008 15:16:06 +0530,

Quoted 2 lines · Harsha priya gupta

Is there any way to avoid the timer part? Because i do not want the cpu

to

Quoted 1 line

bother with anything till the hardware has done its part.

But your hardware doesn't do what we need. A basic part is missing.

Quoted 3 lines

after the hardware
interrupts, can i manipulate someway to get the next buffer rather than interrupting every 1 sec to say that the period has been processed.

Well, how would you handle "the next buffer"?

Suppose all samples on the buffer are processed and an irq is issued. At this moment, you have no data on the buffer. It means that the device is already in a buffer-underrun error state.

The period-based transfer is the basis of ALSA PCM transfer model. In this way, you can write programs requiring a precise timing control, too.

Takashi

Quoted 7 lines · Takashi Iwai

At Mon, 2 Jun 2008 15:06:03 +0530,
Harsha priya gupta wrote:
>
> Say if my hardware is such that it shall interrupt only after it

has

Quoted 2 lines

processed
> entire sample and not ever period or sample. What will ensure that

i get

Quoted 14 lines

my
> next buffer down? Will calling the snd_pcm_period_elapsed in the interrupt
> function help?

So, your hardware has only a single ring buffer and can issue an interrupt only at the end of the buffer?

If so, you might need to seek for another interrupt source, such as a system timer.

Takashi

> On Mon, Jun 2, 2008 at 2:54 PM, Takashi Iwai <[email protected]>

Quoted 7 lines · wrote:
Quoted 7 lines

At Mon, 2 Jun 2008 14:33:14 +0530,
Harsha priya gupta wrote:
>
> Quick question
>
> From my copy function after I pass the buffer to HW, what

would

Quoted 8 lines

happen if
> i
> > call snd_pcm_period_elapsed.
>
> It's invalid and a misdesign.
>
> I guess you are misunderstanding about when to
> callsnd_pcm_period_elapsed(). snd_pcm_period_elapsed() is

called

Quoted 2 lines

when
> one period of samples on the hardware is *processed*. It

doesn't

Quoted 4 lines

mean
> that the samples are transferred to the hardware.
>
> Suppose that you have period_size = 48000 (frames) for 48kHz

samples.

Quoted 1 line
Quoted 1 line

Then, the first snd_pcm_period_epased() shall be called just

one

Quoted 1 line
Quoted 1 line

second after starting the PCM stream. The second call be

another one

Quoted 1 line
Quoted 1 line

second later, and so on. It doesn't matter how quick the copy

to h/w

Quoted 6 lines
Quoted 6 lines
    is done (via copy callback).

    Takashi

    >
    > On Mon, Jun 2, 2008 at 1:37 PM, Takashi Iwai <[email protected]>
Quoted 5 lines · wrote:
Quoted 5 lines
Quoted 5 lines

At Mon, 2 Jun 2008 13:26:01 +0530,
Harsha priya gupta wrote:
>
> I implemented the copy function and immediately

transfered

Quoted 7 lines

the user
> block
> > data
> > > to the hardware.
> > >
> > > Correct me if am wrong;
> > > .pointer implementation - passes the current buffer

pointer.

Quoted 4 lines

When
> the
> > .pointer
> > > function returns the size of the buffer = user buffer

size

Quoted 3 lines
Quoted 3 lines

logically I
> need to
> > expect the hardware to send an interrupt because buffer

is

Quoted 5 lines

consumed
> and I
> > > should call snd_pcm_period_elapsed after that.
> > >
> > > what would happen if i call the snd_pcm_period_elapsed

from

Quoted 3 lines

the
> pointer
> > > function once the buffer is consumed from hardware.

Would

Quoted 7 lines

that be
> right?
> > This
> > > is what i am trying to do
> >
> > The logic is reversed.
> > The pointer callback is a passive one that does nothing

but

Quoted 2 lines

returning
> > the current h/w buffer position. This is called either

from

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

snd_pcm_period_elapsed() or at the PCM status update.

You must call snd_pcm_period_elapsed() somewhere in your

driver

Quoted 8 lines
Quoted 1 line
Quoted 1 line

*explicitly* at the timing that one period is finished.

Usually,
> this
> > is done in an IRQ handler the h/w generates at the period
> ("fragment",
> > "half-buffer", or whatever) boundary.
> >
> > And note that the valid value from the pointer callback

is

Quoted 2 lines

between 0
> > and buffer_size-1 as it handles the buffer as a

ring-buffer.

Quoted 6 lines
     The
    >     >     value buffer_size is invalid.
    >     >
    >     >     Takashi
    >     >
    >     >     > On Mon, Jun 2, 2008 at 1:02 PM, Takashi Iwai <

[email protected]>

Quoted 6 lines

wrote:
> > >
> > > At Mon, 2 Jun 2008 12:39:31 +0530,
> > > Harsha priya gupta wrote:
> > > >
> > > > Can anyone give me a clue as to when i would get

such

Quoted 4 lines

an
> error?
> > >
> > > ... only if you give more clue what exactly you

did.

Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines

In general, it implies that an interrupt isn't

issued

Quoted 25 lines

properly
> at PCM
> > > period boundary.
> > >
> > > Takashi
> > >
> > > --
> > > -Harsha
> > >
> > >
> >
> > --
> > -Harsha
> >
> >
>
> --
> -Harsha
>
>

--
-Harsha

Signature · 1 line
-Harsha
#11 lore raw

At Mon, 2 Jun 2008 15:39:07 +0530,

Quoted 4 lines · Harsha priya gupta

OK. So can i assume that its compulsory to raise at least a timer interrupt every 1 second (for 48K sample/ 48 kHz). if its 96Khz, then i would have to raise the timer interrupt every 2 seconds... correct?

No, no. The one second timer is just my example where the period_size = 48000 at 48kHz is assumed. The actual timer period must be calculated from the sample rate, the buffer size and the period size given by the user and the hardware supports.

Takashi

Quoted 203 lines · Takashi Iwai

At Mon, 2 Jun 2008 15:16:06 +0530,
Harsha priya gupta wrote:
>
> Is there any way to avoid the timer part? Because i do not want the cpu to
> bother with anything till the hardware has done its part.

But your hardware doesn't do what we need. A basic part is missing.

> after the hardware
> interrupts, can i manipulate someway to get the next buffer rather than > interrupting every 1 sec to say that the period has been processed.

Well, how would you handle "the next buffer"?

Suppose all samples on the buffer are processed and an irq is issued. At this moment, you have no data on the buffer. It means that the device is already in a buffer-underrun error state.

The period-based transfer is the basis of ALSA PCM transfer model. In this way, you can write programs requiring a precise timing control, too.

Takashi

>
> On Mon, Jun 2, 2008 at 3:10 PM, Takashi Iwai <[email protected]> wrote:
>
> At Mon, 2 Jun 2008 15:06:03 +0530,
> Harsha priya gupta wrote:
> >
> > Say if my hardware is such that it shall interrupt only after it
has
> processed
> > entire sample and not ever period or sample. What will ensure that
i get
> my
> > next buffer down? Will calling the snd_pcm_period_elapsed in the
> interrupt
> > function help?
>
> So, your hardware has only a single ring buffer and can issue an
> interrupt only at the end of the buffer?
>
> If so, you might need to seek for another interrupt source, such as a
> system timer.
>
> Takashi
>
> > On Mon, Jun 2, 2008 at 2:54 PM, Takashi Iwai <[email protected]> wrote:
> >
> > At Mon, 2 Jun 2008 14:33:14 +0530,
> > Harsha priya gupta wrote:
> > >
> > > Quick question
> > >
> > > From my copy function after I pass the buffer to HW, what
would
> happen if
> > i
> > > call snd_pcm_period_elapsed.
> >
> > It's invalid and a misdesign.
> >
> > I guess you are misunderstanding about when to
> > callsnd_pcm_period_elapsed(). snd_pcm_period_elapsed() is
called
> when
> > one period of samples on the hardware is *processed*. It
doesn't
> mean
> > that the samples are transferred to the hardware.
> >
> > Suppose that you have period_size = 48000 (frames) for 48kHz
samples.
> > Then, the first snd_pcm_period_epased() shall be called just
one
> > second after starting the PCM stream. The second call be
another one
> > second later, and so on. It doesn't matter how quick the copy
to h/w
> > is done (via copy callback).
> >
> > Takashi
> >
> > >
> > > On Mon, Jun 2, 2008 at 1:37 PM, Takashi Iwai <[email protected]>
wrote:
> > >
> > > At Mon, 2 Jun 2008 13:26:01 +0530,
> > > Harsha priya gupta wrote:
> > > >
> > > > I implemented the copy function and immediately
transfered
> the user
> > block
> > > data
> > > > to the hardware.
> > > >
> > > > Correct me if am wrong;
> > > > .pointer implementation - passes the current buffer
pointer.
> When
> > the
> > > .pointer
> > > > function returns the size of the buffer = user buffer
size
> > logically I
> > > need to
> > > > expect the hardware to send an interrupt because buffer
is
> consumed
> > and I
> > > > should call snd_pcm_period_elapsed after that.
> > > >
> > > > what would happen if i call the snd_pcm_period_elapsed
from
> the
> > pointer
> > > > function once the buffer is consumed from hardware.
Would
> that be
> > right?
> > > This
> > > > is what i am trying to do
> > >
> > > The logic is reversed.
> > > The pointer callback is a passive one that does nothing
but
> returning
> > > the current h/w buffer position. This is called either
from
> > > snd_pcm_period_elapsed() or at the PCM status update.
> > >
> > > You must call snd_pcm_period_elapsed() somewhere in your
driver
> > > *explicitly* at the timing that one period is finished.
> Usually,
> > this
> > > is done in an IRQ handler the h/w generates at the period
> > ("fragment",
> > > "half-buffer", or whatever) boundary.
> > >
> > > And note that the valid value from the pointer callback
is
> between 0
> > > and buffer_size-1 as it handles the buffer as a
ring-buffer.
> The
> > > value buffer_size is invalid.
> > >
> > > Takashi
> > >
> > > > On Mon, Jun 2, 2008 at 1:02 PM, Takashi Iwai <
[email protected]>
> wrote:
> > > >
> > > > At Mon, 2 Jun 2008 12:39:31 +0530,
> > > > Harsha priya gupta wrote:
> > > > >
> > > > > Can anyone give me a clue as to when i would get
such
> an
> > error?
> > > >
> > > > ... only if you give more clue what exactly you
did.
> > > >
> > > > In general, it implies that an interrupt isn't
issued
> properly
> > at PCM
> > > > period boundary.
> > > >
> > > > Takashi
> > > >
> > > > --
> > > > -Harsha
> > > >
> > > >
> > >
> > > --
> > > -Harsha
> > >
> > >
> >
> > --
> > -Harsha
> >
> >
>
> --
> -Harsha
>
>

--
-Harsha