Remove extra _deselect to prevent possible hard fault

[jarlamsa]

Description

If read timeout happens, the _deselect will get called twice causing a hard fault happening when mutex is released without being locked.
The SDBlockDevice::read is calling the _deselect in every case.

Pull request type

[x] Fix
[ ] Refactor
[ ] Target update
[ ] Functionality change
[ ] Docs update
[ ] Test update
[ ] Breaking change

[jarlamsa]

@ARMmbed/mbed-os-storage please review

[dannybenor]

@deepikabhavnani Can you help on this review?

[0xc0170]

If read timeout happens, the _deselect will get called twice causing a hard fault happening when mutex is released without being locked.
The SDBlockDevice::read is calling the _deselect in every case.

@jarlamsa This should be part of the commit message to explain why we are removing it, please update

[jarlamsa]

@0xc0170 Do we have some length constraint for the commit message?

[kjbracey]

Just stick to the usual open-source commit message guidelines:

https://tbaggery.com/2008/04/19/a-note-about-git-commit-messages.html

Summary should be short, body can be as many paragraphs as you like.

[jarlamsa]

Thanks @kjbracey-arm, updated

[deepikabhavnani]

If read timeout happens, the _deselect will get called twice

_deselect is called once here https://github.com/ARMmbed/mbed-os/pull/8587/files#diff-fab2d13deb54b56303334ba4f340d4deL895 as per the code change, which is the other location it is called from ?

Never mind, I got confused between _read and _read_bytes earlier. Got the answer _read does not need _deselect on any return, it is taken care of here

mbed-os/components/storage/blockdevice/COMPONENT_SD/SDBlockDevice.cpp

Line 545 in 36f1fa0

_deselect();

[deepikabhavnani]

Good catch.. Looks good to me 👍

[cmonr]

Note: This PR is now a part of a rollup PR (#8663).

No further work is needed here, as once that PR is merged, this PR will also be closed and marked as merged.

If any more commits are made in this PR, this PR will remain open and have to go through CI on its own.