Don't generate .depend anymore

[xiaoxiang781216]

Signed-off-by: Xiang Xiao [email protected]

[patacongo]

@xiaoxiang781216 We have a problem. All chackes are failing because the sim tests are not building. I have the same issue on the PRs that I just submitted. And on Miguel Herranz PR. I merged Miguel's anyway because his ESP32 changes have absolutely nothing to do with sim.

By the way, I requested you review of my changes. That is the removal of CONFIG_FS_WRITABLE that we discussed last week. PR #602

[Ouss4]

@xiaoxiang781216 we might want to get the quiet part back for this: https://github.com/apache/incubator-nuttx/blob/master/tools/sethost.sh#L214
The checks output is unreadable.

[Ouss4]

@xiaoxiang781216 I pulled both this change and the PR from apps to verify that the issue is not similar to the one described in PR602
However the problem persisted. I can build other configs (even custom boards) but the boards failing the checks don't build.
In another note, when one of the failing boards is configured, running distclean results in an infinite loop of :

make[2]: Entering directory '/home/ouss4/NuttX/nuttx/graphics/nxglib'
make[2]: Nothing to be done for 'all'.
make[2]: Leaving directory '/home/ouss4/NuttX/nuttx/graphics/nxglib'

[xiaoxiang781216]

@xiaoxiang781216 I pulled both this change and the PR from apps to verify that the issue is not similar to the one described in PR602
However the problem persisted. I can build other configs (even custom boards) but the boards failing the checks don't build.
In another note, when one of the failing boards is configured, running distclean results in an infinite loop of :

make[2]: Entering directory '/home/ouss4/NuttX/nuttx/graphics/nxglib'
make[2]: Nothing to be done for 'all'.
make[2]: Leaving directory '/home/ouss4/NuttX/nuttx/graphics/nxglib'

My PR make the fail at graphics/nxglib, I will take time to look at the root cause.

[patacongo]

This is a companion to apps/ PR132. Both should be merged together when this one passes its PR checks.

[btashton]

@xiaoxiang781216 @patacongo This seems to have done something to break a builds. I see the CI is passing here but I am not quite clear why. I see lots of these in the logs
/bin/sh: 1: /__w/incubator-nuttx/incubator-nuttx/nuttx/tools/mkdeps: not found and I also see some issues locally where it wont build because of issues running kconf tools in a totally clean environment. When I revert the .depend patch everything goes back to normal.

[Ouss4]

@btashton I'm having issues with this as well. By pulling all the changes from apps/ and also the yet to be merged PR136 a lot of the issues have been resolved.
As for mkdeps, building it manually removes those "not found" messages.
@xiaoxiang781216 one weird issue is that make distclean runs cleanly the first time but generates alot of errors the second time (when everything is clean and no config)
The errors are related to this line https://github.com/apache/incubator-nuttx/blob/master/crypto/Makefile#L96 in all folders with similar Makefile

[yamt]

let's see if a revert fixes the CI breakage. #610

[xiaoxiang781216]

@xiaoxiang781216 @patacongo This seems to have done something to break a builds. I see the CI is passing here but I am not quite clear why. I see lots of these in the logs
/bin/sh: 1: /__w/incubator-nuttx/incubator-nuttx/nuttx/tools/mkdeps: not found and I also see some issues locally where it wont build because of issues running kconf tools in a totally clean environment. When I revert the .depend patch everything goes back to normal.

I am wondering why this patch can pass ci?

[xiaoxiang781216]

let's see if a revert fixes the CI breakage. #610

@yamt the real PR make the break is this one:
apache/nuttx-apps#132
But it's also pass the ci test:(

[Ouss4]

Now, with reverting, the build is broken.
Other PRs (in apps mostly) were merged after this PR (and its companion in apps).