Wrong IIO device attribute (and a iio_rwdev fix) #1237
Conversation
When dumping device attributes, 'waiting_for_supplier' is being included and that is part of device links and has nothing to do with IIO. Signed-off-by: Nuno Sá <[email protected]>
Some minor cosmetics to make the style more consistent. Changes include: * New line between variables declaration and actual code; * Remove some redundant 'else' and 'else if()'; * No new line between variables declaration; * First check for errors instead. Signed-off-by: Nuno Sá <[email protected]>
If we try to cancel the utility with a signal, we'll cancel the buffer (at least on linux) which means we'll likely fail to dequeue a block if we are waiting for one. On top of that the flag 'app_running' is set to false which means that the loop condition '(ret && app_running)' is false and so we do not leave it. This all means that we'll proceed with an invalid block leading to a segfault. Hence, always break the loop if iio_stream_get_next_block() fails (which was the previous behavior) but only log it if the application is running. Signed-off-by: Nuno Sá <[email protected]>
nunojsa
requested review from
cristina-suteu,
dNechita and
mhennerich
as code owners
|
I would rather fix all the style issues in one massive commit with a tool - rather than patching things by hand piecemeal (one file at a time). - using git blame will be harder when we need to ignore multple random commits over time... |
|
Hmm, don't really agree. At least me, 99% (if not 100) of the time, I'll just blame one file to see what commit changed some line and I prefer Hardly applies in this case but if for some reason we wanted to revert the changes in just this file, we could easily revert the commit but if we touch multiple files at a time... Finally, I didn't really wanted to go over all the other files or messing around with a tool to change the style everywhere. But I don't think that should be a show stopper for improving things in the sources I was working with. |
PR Description
Fix a device attribute being wrongly displayed and make sure iio_rwdev does not crash if we hit ctrl-c while streaming with a command as
iio_rwdev -u ip:192.168.1.79 -b 1048576 axi-adrv9002-rx-lpc | pv > /dev/nullThere's also a stylistic patch on local.c
PR Type
PR Checklist