Avoids MSVC warnings about possible value truncation.

Avoids MSVC warnings about possible value truncation.

Suggested by Peter Kasting at
https://bugzilla.mozilla.org/show_bug.cgi?id=917571#c16

R=agl@chromium.org,pkasting@chromium.org
BUG=

Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=291717

[wtc, 2014-08-28 20:45:11]

https://codereview.chromium.org/515383002/diff/1/nss/lib/freebl/chacha20/chac...
File nss/lib/freebl/chacha20/chacha20.c (right):

https://codereview.chromium.org/515383002/diff/1/nss/lib/freebl/chacha20/chac...
nss/lib/freebl/chacha20/chacha20.c:82: input[13] = counter >> 32;

Just wanted to confirm: this line doesn't require a PRUint32 cast, right?

[agl, 2014-08-28 20:51:31]

lgtm

https://codereview.chromium.org/515383002/diff/1/nss/lib/freebl/chacha20/chac...
File nss/lib/freebl/chacha20/chacha20.c (right):

https://codereview.chromium.org/515383002/diff/1/nss/lib/freebl/chacha20/chac...
nss/lib/freebl/chacha20/chacha20.c:82: input[13] = counter >> 32;
On 2014/08/28 20:45:11, wtc wrote:
> Just wanted to confirm: this line doesn't require a PRUint32 cast, right?

I might be tempted to put it anyway. I would guess that the compiler is figuring
out that |counter >> 32| is in-range because of the shift, but that seems like
quite a lot of intelligence.

https://codereview.chromium.org/515383002/diff/1/patches/nss-chacha20-poly130...
File patches/nss-chacha20-poly1305.patch (right):

https://codereview.chromium.org/515383002/diff/1/patches/nss-chacha20-poly130...
patches/nss-chacha20-poly1305.patch:147: +    input[13] = counter >> 32;
ditto.

[Peter Kasting, 2014-08-28 21:11:44]

https://codereview.chromium.org/515383002/diff/1/nss/lib/freebl/chacha20/chac...
File nss/lib/freebl/chacha20/chacha20.c (right):

https://codereview.chromium.org/515383002/diff/1/nss/lib/freebl/chacha20/chac...
nss/lib/freebl/chacha20/chacha20.c:82: input[13] = counter >> 32;
On 2014/08/28 20:51:30, agl wrote:
> On 2014/08/28 20:45:11, wtc wrote:
> > Just wanted to confirm: this line doesn't require a PRUint32 cast, right?
> 
> I might be tempted to put it anyway. I would guess that the compiler is
figuring
> out that |counter >> 32| is in-range because of the shift, but that seems like
> quite a lot of intelligence.

MSVC doesn't require the cast to avoid triggering a warning here, because as agl
suspects, it understands that ">> n" means the top n bits don't matter and can
be safely truncated.

I have no opinion on whether you should add a cast anyway.  I'm not inserting
such casts in other CLs for other codebases I'm modifying as part of this
change.

[wtc, 2014-08-28 21:17:31]

Please review patch set 2. I added a cast to 'counter >> 32'.

[wtc, 2014-08-28 22:13:48]

The CQ bit was checked by wtc@chromium.org

[commit-bot: I haz the power, 2014-08-28 22:14:16]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/wtc@chromium.org/515383002/20001

[commit-bot: I haz the power, 2014-08-28 22:14:47]

Committed patchset #2 (id:20001) as 291717

[agl, 2014-08-29 00:22:51]

lgtm