Bug#1040189: miscomputes SHA1 sums on IMX53 (armhf)

[Simon McVittie]

On Mon, 03 Jul 2023 at 18:04:39 +0300, Dmitry Baryshkov wrote:
> I wrote this in a 'bts close' email, but it seems bts hides this
> somewhere. I didn't notice that default settings overclock the board.
> And most of the things work, except the rare fallouts (like sha1).
> After limiting the CPU frequency to 1GHz (from the default 1.2 GHz)
> the system is stable (and all hashsums match).

I'm glad you found a solution! It's unfortunate if this hardware is
being sold in a state where it doesn't work correctly...

For your information, `bts close` is undocumented (perhaps intentionally?)
and doesn't seem to notify the package's maintainer or submitter. If
in doubt, the right thing is to reply to the bug address (NNNNNN@
where NNNNNN is the bug number) with human-readable information, and
the package's maintainer can sort out the BTS state.

The usual way to close a bug is to send mail to NNNNNN-done@ where
NNNNNN is the bug number. The message body should start with something
like "Version: 1.2-3" followed by a blank line if the bug was fixed by a
version of the package, or should not specify a version in cases like this
one where the bug is being closed as "not a bug", "no longer applicable",
"fixed elsewhere" or similar. I believe `bts done` is a shortcut for
this, but I've never used that particular feature since closing a bug
requires a human-readable explanation anyway.

    smcv