#7230 Internal cardreader NEEDINIT (no entitlements)

[npj0808]

Revision

<7230>

Issue Description

<no entitlements on et9000>

When the issue occurs

<after 7229>

Sorry i have no log file again !!

[theparasol]

We got this ticket before the same answer as before...: Clarketech needs to change their internal readercode.

Ok 1 last tryfix after reading ​http://www.www.streamboard.tv:8001/ticket/2643# again

[npj0808]

Sorry fix are not working !! and what i want to ask an Clarktech??

I become also no ATR !!!! anymore !!

[theparasol]

This is duplicate and no forum ​http://www.www.streamboard.tv:8001/ticket/2643#comment:14

[ajeeb]

yes also 7231 !!! I go back to 6xxxx ! plz fix that

[Deas]

can you read? this is a driver issue. so use openpli 3.0 image as posted by kalehrl or open a bug report with the vendor. this is not an oscam issue. and no stop replying in closed tickets. if you want to discuss this issue open a thread in the forum!

[littlesat]

This is not a driver issue... This is a Oscam bug...

Not having looked at the code, only browsing through the discussion in the ticket, the solution when EAGAIN occurs seems to be to go back into the poll/select loop, instead of having a tight loop around the read.

EAGAIN = EWOULDBLOCK, which indicates "I have no data, I would block if I were in blocking mode".

And "no data" could of course still be "no data" after an infinite number of retries.

Going back to the poll/select would probably show the filedescriptor is no longer readable.

Note that EAGAIN is not the same as EINTR.

EINTR requires just the interrupted call to be restarted.
For EAGAIN, when it occurs in a read, it would not be a very good idea to just keep restarting the system call.