summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorDavid Young <dyoung@hdfgroup.org>2019-11-25 17:01:42 (GMT)
committerDavid Young <dyoung@hdfgroup.org>2019-11-25 17:01:42 (GMT)
commit4d50f9a7a8737380b37b7cde751dbeae3b7c6b7c (patch)
tree0d17e2f9133bb4536fb7084d45d73201da872f94
parent0f9678b41526cf935506f11e389b5ccec3519f5b (diff)
downloadhdf5-4d50f9a7a8737380b37b7cde751dbeae3b7c6b7c.zip
hdf5-4d50f9a7a8737380b37b7cde751dbeae3b7c6b7c.tar.gz
hdf5-4d50f9a7a8737380b37b7cde751dbeae3b7c6b7c.tar.bz2
Update status of the expand/shrink test.
-rw-r--r--doc/VFD_SWMR_Punch_List.md13
1 files changed, 7 insertions, 6 deletions
diff --git a/doc/VFD_SWMR_Punch_List.md b/doc/VFD_SWMR_Punch_List.md
index 5a80995..5970bd4 100644
--- a/doc/VFD_SWMR_Punch_List.md
+++ b/doc/VFD_SWMR_Punch_List.md
@@ -120,15 +120,16 @@ guide to the syntax.
SWMR does not have to deal with it. However, if the global heap
overhaul does not take place, then we have more work to do.
-26. **David, in-progress** Fix the expand/shrink test.
+26. **David, needs merge** Fix the expand/shrink test.
- The test *probably* fails because the dataset extent is enlarged
- before the data is written, so arbitrary data is present until the
- data appears. If a tick sneaks in between the `H5Dset_extent` and
- the `H5Dwrite`, then the reader may read the arbitrary data.
+ The test appeared to fail because the dataset extent was enlarged
+ before the data was written, so arbitrary data was present.
+ If a tick snuck in between the `H5Dset_extent` and
+ the `H5Dwrite`, then the reader read the arbitrary data.
In the `gaussian` test, I have a heuristic that avoids reading
- arbitrary data. I may replicate that in the expand/shrink test.
+ arbitrary data. Replicating that in the expand/shrink test has
+ fixed it. Essentially, the reader trails the writer by a bit.
Ultimately, we should suspend ticks over the H5Dset_extent/H5Dwrite.