[sane-devel] deadlock with sanei_usb_read_bulk
excellent work. please keep us informed.
allan
On 2/10/08, J?rgen Ernst jrernst at gmx.de wrote:
Hi!
I managed to solve the problem.
Here it is:
Canon driver sends a dummy bulk write as shown in the log. It only
produces errors and I think it's only for delaying bulk reads. We can
omit that and use a fix delay of 1 ms instead.
Sanei_USB sets the right endpoints. I can read and write data on
endpoint 0. It's all as it should be. Now initializing scanner and
reading buttons is working. Getting access to scandata is following.
J?rgen Ernst wrote:
Hi!
I'm trying to write the backend for Canon LiDE 600F and now I'm stuck
with sanei_usb_read_bulk. I tried 10 hours but I don't understand how to
get sanei_usb working with bulk_read. Can somebody help me out?
Here are some details:
Canon LiDE 600F is similar to Canon LiDE 70. I checked this with two
guys owning a LiDE 70. So you can read some info on
http://www.sane-project.org/unsupported/canon-lide-70.html
...
endpoint 0
bEndpointAddress 0x02 (out 0x02)
bmAttributes 2 (bulk)
...
endpoint 0
bEndpointAddress 0x83 (in 0x03)
bmAttributes 2 (bulk)
...
For short:
LiDE 600F has only two endpoints. Both bulk no interrupt. In the log
file no control messages are used. Only simple bulk read and write.
I programmed a perl script with which I was able to access the scanner
and retrieve scan data.
So this was the point to start writing the sane backend in C.
Okay, let's look first on a perl code snippet:
sub canon_get
{
my ( $reg ) = @_ ;
my $data = '01 ' . $reg . ' 01 00' ;
my $err = $dev-bulk_write(hex('02'), hex2bin($data), $timeout);
my $data = \0 x 512 ;
my $cnt = $dev-bulk_write(hex('03'), $data, $timeout);
my $data = ' ' ;
my $cnt = $dev-bulk_read(hex('83'), $data, 1, $timeout) ;
return ( $data ) ;
}
This is the procedure to get 1 byte of data from the CP2155 chip.
Suppose we called canon_get('91'); then $reg is '91' in hex.
$data is set to '01 91 01 00'.
I made it this way to better check with output from log.
hex2bin converts this string to binary.
Then a first bulk_write is made. In perl I had to give the
bEndpointAddress '02' while in sanei_usb I saw I had to give the
endpoint number as an integer (endpoint 0).
Second is a dummy write on the other endpoint (endpoint 1) with 512
bytes of data.
Third we can get the result with a 1 byte bulk_read on endpoint 1.
That's it.
But it doesn't work in sane.
Here's the snippet of my test procedure in C:
static byte setup_buffer[] = { 0, 0, 0, 0, ... ,0 };
static SANE_Status
CP2155get (CP2155_Register reg, byte *data)
{
SANE_Status status;
size_t count;
setup_buffer[0] = 0x01;
setup_buffer[1] = (reg) 0xFF;
setup_buffer[2] = 0x01;
setup_buffer[3] = 0x00;
count = 4;
status = sanei_usb_write_bulk (0, setup_buffer, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_write_bulk.1 error\n);
count = 512;
status = sanei_usb_write_bulk (0, setup_buffer, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_write_bulk.2 error\n);
count = 1;
status = sanei_usb_read_bulk (0, data, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_read_bulk error\n);
return status;
}
setup_buffer is an array with 512 bytes.
The first two bulk_writes didn't return any error.
It seems that they worked fine.
On the last bulk_read the program hangs forever.
I surely did a mistake but I don't know where and why.
Please help.
--
sane-devel mailing list: sane-devel at lists.alioth.debian.org
http://lists.alioth.debian.org/mailman/listinfo/sane-devel
Unsubscribe: Send mail with subject unsubscribe your_password
to sane-devel-request at lists.alioth.debian.org
--
The truth is an offense, but not a sin
[sane-devel] deadlock with sanei_usb_read_bulk
Hi!
I'm trying to write the backend for Canon LiDE 600F and now I'm stuck
with sanei_usb_read_bulk. I tried 10 hours but I don't understand how to
get sanei_usb working with bulk_read. Can somebody help me out?
Here are some details:
Canon LiDE 600F is similar to Canon LiDE 70. I checked this with two
guys owning a LiDE 70. So you can read some info on
http://www.sane-project.org/unsupported/canon-lide-70.html
...
endpoint 0
bEndpointAddress 0x02 (out 0x02)
bmAttributes 2 (bulk)
...
endpoint 0
bEndpointAddress 0x83 (in 0x03)
bmAttributes 2 (bulk)
...
For short:
LiDE 600F has only two endpoints. Both bulk no interrupt. In the log
file no control messages are used. Only simple bulk read and write.
I programmed a perl script with which I was able to access the scanner
and retrieve scan data.
So this was the point to start writing the sane backend in C.
Okay, let's look first on a perl code snippet:
sub canon_get
{
my ( $reg ) = @_ ;
my $data = '01 ' . $reg . ' 01 00' ;
my $err = $dev-bulk_write(hex('02'), hex2bin($data), $timeout);
my $data = \0 x 512 ;
my $cnt = $dev-bulk_write(hex('03'), $data, $timeout);
my $data = ' ' ;
my $cnt = $dev-bulk_read(hex('83'), $data, 1, $timeout) ;
return ( $data ) ;
}
This is the procedure to get 1 byte of data from the CP2155 chip.
Suppose we called canon_get('91'); then $reg is '91' in hex.
$data is set to '01 91 01 00'.
I made it this way to better check with output from log.
hex2bin converts this string to binary.
Then a first bulk_write is made. In perl I had to give the
bEndpointAddress '02' while in sanei_usb I saw I had to give the
endpoint number as an integer (endpoint 0).
Second is a dummy write on the other endpoint (endpoint 1) with 512
bytes of data.
Third we can get the result with a 1 byte bulk_read on endpoint 1.
That's it.
But it doesn't work in sane.
Here's the snippet of my test procedure in C:
static byte setup_buffer[] = { 0, 0, 0, 0, ... ,0 };
static SANE_Status
CP2155get (CP2155_Register reg, byte *data)
{
SANE_Status status;
size_t count;
setup_buffer[0] = 0x01;
setup_buffer[1] = (reg) 0xFF;
setup_buffer[2] = 0x01;
setup_buffer[3] = 0x00;
count = 4;
status = sanei_usb_write_bulk (0, setup_buffer, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_write_bulk.1 error\n);
count = 512;
status = sanei_usb_write_bulk (0, setup_buffer, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_write_bulk.2 error\n);
count = 1;
status = sanei_usb_read_bulk (0, data, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_read_bulk error\n);
return status;
}
setup_buffer is an array with 512 bytes.
The first two bulk_writes didn't return any error.
It seems that they worked fine.
On the last bulk_read the program hangs forever.
I surely did a mistake but I don't know where and why.
Please help.
--
Ciao
J?rgen
Microsoft is not the answer. Microsoft is the question.
NO is the answer.
[sane-devel] deadlock with sanei_usb_read_bulk
Hi!
I managed to solve the problem.
Here it is:
Canon driver sends a dummy bulk write as shown in the log. It only
produces errors and I think it's only for delaying bulk reads. We can
omit that and use a fix delay of 1 ms instead.
Sanei_USB sets the right endpoints. I can read and write data on
endpoint 0. It's all as it should be. Now initializing scanner and
reading buttons is working. Getting access to scandata is following.
J?rgen Ernst wrote:
Hi!
I'm trying to write the backend for Canon LiDE 600F and now I'm stuck
with sanei_usb_read_bulk. I tried 10 hours but I don't understand how to
get sanei_usb working with bulk_read. Can somebody help me out?
Here are some details:
Canon LiDE 600F is similar to Canon LiDE 70. I checked this with two
guys owning a LiDE 70. So you can read some info on
http://www.sane-project.org/unsupported/canon-lide-70.html
...
endpoint 0
bEndpointAddress 0x02 (out 0x02)
bmAttributes 2 (bulk)
...
endpoint 0
bEndpointAddress 0x83 (in 0x03)
bmAttributes 2 (bulk)
...
For short:
LiDE 600F has only two endpoints. Both bulk no interrupt. In the log
file no control messages are used. Only simple bulk read and write.
I programmed a perl script with which I was able to access the scanner
and retrieve scan data.
So this was the point to start writing the sane backend in C.
Okay, let's look first on a perl code snippet:
sub canon_get
{
my ( $reg ) = @_ ;
my $data = '01 ' . $reg . ' 01 00' ;
my $err = $dev-bulk_write(hex('02'), hex2bin($data), $timeout);
my $data = \0 x 512 ;
my $cnt = $dev-bulk_write(hex('03'), $data, $timeout);
my $data = ' ' ;
my $cnt = $dev-bulk_read(hex('83'), $data, 1, $timeout) ;
return ( $data ) ;
}
This is the procedure to get 1 byte of data from the CP2155 chip.
Suppose we called canon_get('91'); then $reg is '91' in hex.
$data is set to '01 91 01 00'.
I made it this way to better check with output from log.
hex2bin converts this string to binary.
Then a first bulk_write is made. In perl I had to give the
bEndpointAddress '02' while in sanei_usb I saw I had to give the
endpoint number as an integer (endpoint 0).
Second is a dummy write on the other endpoint (endpoint 1) with 512
bytes of data.
Third we can get the result with a 1 byte bulk_read on endpoint 1.
That's it.
But it doesn't work in sane.
Here's the snippet of my test procedure in C:
static byte setup_buffer[] = { 0, 0, 0, 0, ... ,0 };
static SANE_Status
CP2155get (CP2155_Register reg, byte *data)
{
SANE_Status status;
size_t count;
setup_buffer[0] = 0x01;
setup_buffer[1] = (reg) 0xFF;
setup_buffer[2] = 0x01;
setup_buffer[3] = 0x00;
count = 4;
status = sanei_usb_write_bulk (0, setup_buffer, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_write_bulk.1 error\n);
count = 512;
status = sanei_usb_write_bulk (0, setup_buffer, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_write_bulk.2 error\n);
count = 1;
status = sanei_usb_read_bulk (0, data, count);
if (status != SANE_STATUS_GOOD)
DBG (1, CP2155get: sanei_usb_read_bulk error\n);
return status;
}
setup_buffer is an array with 512 bytes.
The first two bulk_writes didn't return any error.
It seems that they worked fine.
On the last bulk_read the program hangs forever.
I surely did a mistake but I don't know where and why.
Please help.
