Re: [Qemu-devel] [V4 PATCH 05/22] softfloat: Fix float64_to_uint32_round_to_zero

[Peter Maydell]

On 18 December 2013 20:19, Tom Musta <address@hidden> wrote:
> The float64_to_uint32_round_to_zero routine is incorrect.
>
> For example, the following test pattern:
>
>     425F81378DC0CD1F / 0x1.f81378dc0cd1fp+38
>
> will erroneously set the inexact flag.
>
> This patch re-implements the routine to use the 
> float64_to_uint64_round_to_zero
> routine.  If saturation occurs and the inexact flag was not previously set, it
> will be cleared.

Code is OK but you forgot to update this commit message.

>
> This contribution can be licensed under either the softfloat-2a or -2b
> license.
>
> V4: Correct commit commentary.  Corrected code to properly handle the
> stickiness of the inexact flag.

These 'changes from previous version' comments should ideally go
below the "---" line, by the way. The idea is that the bit above the
line goes into the git commit history and should be a self-contained
description of the final patch.

> Signed-off-by: Tom Musta <address@hidden>

If you fix the commit message you can add:
Reviewed-by: Peter Maydell <address@hidden>

thanks
-- PMM