Static analysis

[Dimitriy Ryazantcev]

Checked code with PVS Studio static analyzer. Found bunch of minor(?) errors in bitwise operations etc.

Consider to check the list attached please. There are many false positive through it but I think some messages might be useful.

[Dimitriy Ryazantcev]

Explanation for this VXXX "warnings" you can fnind there: http://www.viva64.com/en/d/

[Steve Borho]

Thanks Dmytro,

Do you happen to remember what version of the source code you ran the static analyzer on? Most of these reported line numbers are no longer usable.

[Matteo Gazzetta]

hope it helps

[Dimitriy Ryazantcev]

Analysing was done on 25dde1f revision.

[Deepthi Nandakumar]

entropy: fix SAO enable detection (refs #80)

Apparently our analysis never toggles luma separately from chroma because this bug has not resulted in any bad bitstreams, that I know of. This bug was found via static analysis

→ <<cset c39538f0c59b>>

[Deepthi Nandakumar]

encoder: use %u to sprintf unsigned ints (refs #80)

→ <<cset a58aea624122>>

[Deepthi Nandakumar]

TComWeightPrediction: combine duplicate inline functions (refs #80)

→ <<cset c7cc07fd21a7>>

[Dimitriy Ryazantcev]

Created pull request #16 with some fixes based on this analysis.

[Steve Borho]

any chance you can re-run with a newer revision?

[Dimitriy Ryazantcev]

Sure Steve! From rev.4495af3b30bb

[Matteo Gazzetta]

From rev. 2b93cf2

[Steve Borho]

y4m: use ios::good() rather than !ios::eof() (refs #80)

This is more robust to file read failures and will prevent infinite loops in the event of a failure without EOF. good() imples !fail && !eof && !bad

→ <<cset bfc5a2e98f86>>

[Steve Borho]

not much point in keeping this issue open