add possibility to configure random stalls for axi_stream #557
Conversation
|
I am ok with this. But there are merge conflicts. @LarsAsplund @1138-4eb are you ok with this? |
|
I don't understand why the tests are failing, they run without failing on my local machine... |
|
Ok, somehow the last changes got stuck along the way... |
Which simulator are you using? Tests run with GHDL in a docker container. You can run them locally with: docker run --rm \
-v /$(pwd)://src \
-w //src \
vunit/dev:llvm \
tox -e py38-vcomponents-ghdl |
| constant slave_stream : stream_slave_t := as_stream(slave_axi_stream); | ||
| constant slave_sync : sync_handle_t := as_sync(slave_axi_stream); | ||
|
|
||
| constant monitor : axi_stream_monitor_t := new_axi_stream_monitor( | ||
| data_length => 8, id_length => g_id_length, dest_length => g_dest_length, user_length => g_user_length, | ||
| logger => get_logger("monitor"), actor => new_actor("monitor"), | ||
| data_length => 8, id_length => g_id_length, dest_length => g_dest_length, user_length => g_user_length, |
There was a problem hiding this comment.
The format here is different from slave_axi_stream or master_axi_stream above. I'd personally put each parameter in a line. However, since a different style is already being used, I think that it should be preserved. Specially, the additional spaces in lines 39, 48, 56, 62, 124, 164, etc. and in lines 53, 54, 354, etc. Nevertheless, I think that the alignment fixes in 126-129, 427-434, 457-464, etc. are good.
Alternatively, if you did this beautification automatically with some tool, we'd be glad to know which did you use. We are currently using black for Python, but we don't know of any open source tool that allows to do the same with VHDL.
There was a problem hiding this comment.
I'm not quite sure how to interpret your comment, should I revert beautification?
This is the emacs beautification. I'm actually using neovim but the emacs beautification can be called in batch mode.
IMHO emacs has the only 'correct' indentation for vhdl code.
emacs --no-site-file -batch <FILE> -f vhdl-beautify-buffer -f save-buffer
There was a problem hiding this comment.
I'm not quite sure how to interpret your comment, should I revert beautification?
Yes. It's ok to fix misaligned groups of parameters and other minor issues around your changes; but some of the style changes introduce inconsistencies that should be handled in a separate PR. For example, this specific line feels awkward.
Thanks for pointing to emacs. Although it seems that no new features are to be added (https://guest.iis.ee.ethz.ch/~zimmi/emacs/vhdl-mode.html#maintenance), VHDL 2008 is said to be supported. (Un)fortunately, the style can be customized: https://www.gnu.org/software/emacs/manual/html_mono/vhdl-mode.html#Customizing-Indentation. On the one hand, did you use any specific style configuration? On the other hand, this discussion might deserve a separate issue before a PR.
There was a problem hiding this comment.
No, I did not use any specific style configuration.
There was a problem hiding this comment.
Some minor fixes, but LGTM. @kraigher, @LarsAsplund, can you have a look at the two points above?
|
@dstadelm, I rebased on top of master and I force-pushed the changes to your branch/fork, so that the conflicts are gone and tests are properly executed. |
|
Please review, thx |
|
@dstadelm, it seems that, instead of making your latest changes on top of the rebased branch commented in #557 (comment) (i.e. 6f769f9), you changed your local branch and force-pushed on top. Please, either rebase it yourself or apply latest fixes on top of 6f769f9 . |
|
I had to merge, rebase f***ed up good time. I've never experienced this before, but I just couldn't figure out what went wrong. |
|
@dstadelm, what's the status of this PR? I saw that you reacted to some of Lars' comments, but it seems that you didn't push any commit after that. Did you forget to push it or have you been busy? Just asking to know when to help you with rebasing. |
|
@eine We had Christmas vacation (family time:) and I had a release for our customer. I will work on it now... |
|
@eine |
|
The CI failures seem to be due to some uppercasing I pushed some days ago. Easy to fix. At the same time, docs are failing due to an unrelated issue. I'll try to take care of it before merging this. |
Ignore them. It is unrelated to your changes, and I'm taking care of it: https://github.com/eine/vunit/commits/rm-ablog You should only worry about vcomponents tests. |
|
@dstadelm, you can now rebase, and the docs should pass. |
At creation of master/slave one can provide a configuration which tells how often a stall should occure, and if it occures what the minimal/maximal stall length of such a stall shall be.
based on pull request comment of 1138-4EB
Changed new_actor to p_actor
this makes it more general and percentage numbers can be varied through the generics without having to manual change the vhdl code.
eine
changed the title
|
Thanks @dstadelm! Nice work! 🎉 |
At creation of axi_stream_master/slave one can provide a configuration which tells
how often a stall should occur, and if it occurs what the
minimal/maximal stall length of such a stall shall be.