Re: [Qemu-devel] [PATCH v4 0/2] block: Reject negative values for throttling options

[Markus Armbruster]

Fam Zheng <address@hidden> writes:

> v4: Add Max's rev-by in both patches, while fixing the "maxs" typo.
>
> v3: Address comments:
>     - Add test for large value; [Berto]
>     - Fix typos "negative" & "caught"; [Eric, Berto]
>     - Use "LL" suffix to the upper limit constant. [Berto]
>
> v2: Check the value range and report an appropriate error. [Berto]
>
> Now the negative values are silently converted to a huge positive number
> because we are doing implicit casting from uint64_t to double. Fix it and add 
> a
> test case (this was once fixed in 7d81c1413c9 but regressed when the block
> device option parsing code was changed).

I think PATCH 1's commit message could explain the problem in a bit more
detail, and it should mention the changed valid range.

Other than that, I had two questions: why cast THROTTLE_VALUE_MAX for
printing (in scope for the series), and why parse the settings as
integers even though they're really floating-point (probably not in
scope).