Thank you so much for your support allan. I will take note of this. On Wed, Mar 9, 2016 at 8:17 PM, m. allan noah <kitno...@gmail.com> wrote:
> Yes- the value of maxlen is determined by the frontend. You cannot > change it, and you certainly cannot re-alloc. You should have your own > buffer in the backend, and read bytes from it into the buffer provided > by the frontend. This type of double buffering will allow you to read > from the scanner in whatever block size is best, and then respond to > the frontend however it wants. I recommend that you not make your > backend buffer very large (to save memory), but you then have to be > careful not to overflow it when the frontend is draining it slowly. > > allan > > On Wed, Mar 9, 2016 at 1:14 AM, ky gcp <kygcpsam...@gmail.com> wrote: > > Thank you for such quick response, allan. Do you mean to say that the > value > > of maxlen is determined by the frontend? On xsane application, maxlen > value > > is 8192 when using acquire preview and then value of maxlen is 65536 when > > using scan function. Our external backend currently reads images by row > on > > sane_read(). Issue occurs when using acquire preview on colored A3 > original > > because the number of bytes per row is more than the value of maxlen. If > I > > reallocate the argument buf to a much bigger size, xsane closes and I get > > this error message below: > > *** stack smashing detected ***: xsane terminated > > Aborted (core dumped) > > > > Since our external backend currently reads images by row on sane_read(), > is > > it better to just change our implementation on sane_read() ? > > > > On Wed, Mar 9, 2016 at 9:39 AM, m. allan noah <kitno...@gmail.com> > wrote: > >> > >> Because different frontends have different requirements. Some try to > >> save memory, some try to maximize speed by using large buffers, some > >> try to fill a network packet exactly. You can (and should expect) to > >> get any value of maxlen from 1 byte to perhaps a 1M, maybe more. > >> > >> allan > >> > >> On Tue, Mar 8, 2016 at 8:23 PM, ky gcp <kygcpsam...@gmail.com> wrote: > >> > why does the value of maxlen vary from various frontends? > >> > > >> > On Tue, Mar 8, 2016 at 9:06 PM, m. allan noah <kitno...@gmail.com> > >> > wrote: > >> >> > >> >> maxlen is provided by the caller of the function, to tell the backend > >> >> how much memory the caller has allocated for image data. > >> >> > >> >> allan > >> >> > >> >> On Tue, Mar 8, 2016 at 3:09 AM, ky gcp <kygcpsam...@gmail.com> > wrote: > >> >> > hi, > >> >> > > >> >> > how is the argument maxlen determined in sane_read() function? > >> >> > > >> >> > SANE_Status sane_read (SANE_Handle h, SANE_Byte * buf, SANE_Int > >> >> > maxlen, > >> >> > SANE_Int * len); > >> >> > > >> >> > Thanks for your help. > >> >> > > >> >> > -- > >> >> > sane-devel mailing list: sane-devel@lists.alioth.debian.org > >> >> > http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel > >> >> > Unsubscribe: Send mail with subject "unsubscribe your_password" > >> >> > to sane-devel-requ...@lists.alioth.debian.org > >> >> > >> >> > >> >> > >> >> -- > >> >> "well, I stand up next to a mountain- and I chop it down with the > edge > >> >> of my hand" > >> > > >> > > >> > >> > >> > >> -- > >> "well, I stand up next to a mountain- and I chop it down with the edge > >> of my hand" > > > > > > > > -- > "well, I stand up next to a mountain- and I chop it down with the edge > of my hand" >
-- sane-devel mailing list: sane-devel@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject "unsubscribe your_password" to sane-devel-requ...@lists.alioth.debian.org
