Some symbols not indexed about elixir

[michaelopdenacker]

Hi Chris,
Wow, reindexing the whole history takes time, because our database doesn't scale (we will address this through an internship in April: https://bootlin.com/internships/2020/en/internship-elixir.pdf)
That's done now. See https://elixir.bootlin.com/linux/latest/ident/device
Thanks again,
Michael.

[michaelopdenacker]

Hey Chris,
Thanks in advance, do your best. This will be highly appreciated by a vast number of users :)
Cheers,
Michael.

[michaelopdenacker]

The above comment is not exactly true. "dscr" is defined elsewhere (see https://elixir.bootlin.com/linux/v5.4-rc1/ident/dscr), and hence is recognized as an identifier when we parse the code in https://elixir.bootlin.com/linux/v5.4-rc1/source/arch/arm/kernel/hw_breakpoint.c

[cxw42]

I'm not sure if this is the same issue, but I just noticed that struct device is a bit odd:

• At the definition, device is not linked or highlighted.
  • This despite the fact that other structs in that file are, e.g., bus_type here.
• The search does list struct device in the results.

Let me know if this is unrelated. I did check other issues first but didn't find one that seemed to fit. Thanks!

[michaelopdenacker]

Hi Chris,
Many thanks for this bug report.
Yes, this is unrelated to this bug. This happens because the "device" structure was wrongly blacklisted.
That's fixed now (see cb49526), but it will take a few days to fix, as I'll need to regenerate the database completely.
Thanks again!
Michael

[cxw42]

Michael, much appreciated!

[cxw42]

Michael, looks good! Yes, I left the indexer running on my machine over a weekend and only got 1/3 of the way through! I am glad to hear there are changes in the works. I will try to get my PR for #78 in by the end of March.

[michaelopdenacker]

Closing this bug
We don't want to support local variables.