Mercurial > public > ostc4
diff Discovery/Src/buehlmann.c @ 261:cc2406b835ff bm-3
Bugfix: do not reset saturation on surfacing
Commit 822416168585 introduced a subtle bug. On surfacing, the value of
saturation was reset to 0. This is prefect proof why global data is a
dangerous thing, and subtle changes can introduce seemingly unrelated
bugs.
While it would be much better to factor out as much as possible global
data, the fix here does not do that. Simply, the bug is fixed without
touching the rather complex gTissue_nitrogen_bar/gTissue_helium_bar
handling.
Signed-off-by: Jan Mulder <jlmulder@xs4all.nl>
| author | Jan Mulder <jlmulder@xs4all.nl> |
|---|---|
| date | Tue, 23 Apr 2019 13:05:20 +0200 |
| parents | 0087d6251f59 |
| children | 24ff72e627f4 |
line wrap: on
line diff
--- a/Discovery/Src/buehlmann.c Tue Apr 23 09:25:18 2019 +0200 +++ b/Discovery/Src/buehlmann.c Tue Apr 23 13:05:20 2019 +0200 @@ -630,6 +630,9 @@ { float ceiling; + memcpy(gTissue_nitrogen_bar, pLifeData->tissue_nitrogen_bar, (4*16)); + memcpy(gTissue_helium_bar, pLifeData->tissue_helium_bar, (4*16)); + // this is just performance optimizing. The code below runs just fine // without this. There is never a ceiling in NDL deco state if (!pDecoInfo->output_time_to_surface_seconds) { @@ -637,9 +640,6 @@ return; } - memcpy(gTissue_nitrogen_bar, pLifeData->tissue_nitrogen_bar, (4*16)); - memcpy(gTissue_helium_bar, pLifeData->tissue_helium_bar, (4*16)); - ceiling = compute_ceiling(pLifeData->pressure_surface_bar, 1.0f + pLifeData->max_depth_meter/10.0f); pDecoInfo->output_ceiling_meter = (ceiling - pLifeData->pressure_surface_bar) * 10.0f; }