crash plotting from a silo filo.

[markcmiller86]

Date: Tue, 10 Mar 2020 13:44:42 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

I got the following bug report from a user.

I have a certain material that contains more plot variables being output to a SILO file than other
materials within the same simulation.  If I run the simulation on multiple processes and output to
SILO, then it behaves fine and I am able to view those extra plot variables in Visit for only the
material which has them (see screenshot below, only the colored element has the variable being
plotted).  However, if I run the same simulation on only 1 process, then I get an error when trying
to view this same variable in Visit and it crashes.

![visit_image](https://user-
images.githubusercontent.com/1102718/76357528-1ea66980-62d5-11ea-9dcf-a9270920f854.png)

I have looked through our SILO write routines within Diablo and do not see anything obvious that
would suggest why I would get different behavior when running with multiple processes vs. a single
process. I?ve manually inspected the SILO hdf files using ?hdfview? and don?t see any obvious
differences other than the number of procs.  It may quite possibly still be something on our end,
but was wondering if either of you had seen an issue like this before?  I can provide the relevant
SILO files if that would help.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200310/d65a05b6/attachment.html>

---

[markcmiller86]

Date: Tue, 10 Mar 2020 13:46:26 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

There shouldn?t be any difference in behavior when running in serial or parallel. Can you give me
the silo files? When you created the  pseudocolor plot, did you material select to only the
materials that have the variable or did you do it over the entire mesh?

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-597305992

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200310/15128c13/attachment.html>

---

[markcmiller86]

Date: Tue, 10 Mar 2020 13:47:01 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

I gave you an example of the SILO files created when I run with 2 procs vs 1 proc.  The variable I
plotted below is pseudocolor ?eeyy?.

Regarding the pseudocolor plot, I did not do any special selection of materials. It should be
covering the whole mesh.  In our code it seems that we currently pass a NULL ptr into the SILO
routines for element sets which do not contain the variable, though this should be the same
regardless of num_procs.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-597306254

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200310/bcbbbfd4/attachment.html>

---

[markcmiller86]

Date: Fri, 13 Mar 2020 16:25:25 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

I have a fix that allows creating a pseudocolor plot of 'eeeyy'. The fix involved more changes than
I thought, fortunately in just 3 files. There are 2 more spots that don't affect your data that I
need to fix. I can make a version available now if you need it or in a little bit when I finish all
the changes for you to test.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-598970898

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200313/c6c9e506/attachment.html>

---

[markcmiller86]

Date: Mon, 16 Mar 2020 14:15:12 -0700
From: notifications at github.com (Alister)
This post was imported from the [email protected] email archive

---

@brugger1 should I move this to the main issue tracker?

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-599761547

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200316/e4361e0b/attachment.html>

---

[markcmiller86]

Date: Mon, 16 Mar 2020 14:22:06 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

I'm waiting to hear back from the user.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-599764037

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200316/67df7232/attachment.html>

---

[markcmiller86]

Date: Mon, 16 Mar 2020 15:06:07 -0700
From: notifications at github.com (Mark C. Miller)
This post was imported from the [email protected] email archive

---

> @brugger1 should I move this to the main issue tracker?

Yes. But, did it turn out to be a Silo issue or a VisIt issue? If silo, I can move to the Silo issue
tracker

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-599779049

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200316/db5be001/attachment.html>

---

[markcmiller86]

Date: Mon, 16 Mar 2020 15:11:14 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

It?s a VisIt issue handling creating ghost zones from global ids and some of the blocks being NULL.
I?ll have a pull request with the fix soon that you can look at.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-599780708

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200316/54ee3798/attachment.html>

---

[markcmiller86]

Date: Mon, 30 Mar 2020 08:57:36 -0700
From: notifications at github.com (Mark C. Miller)
This post was imported from the [email protected] email archive

---

@brugger...is this now resolved?

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-606086485

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200330/d77051a6/attachment.html>

---

[markcmiller86]

Date: Mon, 30 Mar 2020 10:29:23 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

The fix is checked in, but I am waiting to hear back from the user. It turns out the e-mail came
directly to you and me and I responded here, erroneously thinking he would get it from visit-users.
I just now responded in the original e-mail chain, so I hope to hear back from the customer soon.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-606135575

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200330/c3fa35bd/attachment.html>

---

[markcmiller86]

Date: Mon, 30 Mar 2020 10:35:07 -0700
From: wangyechen9567 at 163.com (=?UTF-8?B?546L5Y+25pmo?=)
This post was imported from the [email protected] email archive

---

Why i have received so many emails of this kind? It's really bothering me

| |
???
|
|
???wangyechen9567 at 163.com
|

??? ?????? ??

On 03/31/2020 01:29, Eric Brugger via visit-users wrote:

The fix is checked in, but I am waiting to hear back from the user. It turns out the e-mail came
directly to you and me and I responded here, erroneously thinking he would get it from visit-users.
I just now responded in the original e-mail chain, so I hope to hear back from the customer soon.

?
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub, or unsubscribe.

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200331/864695b8/attachment.html>

---

[markcmiller86]

Date: Wed, 20 May 2020 15:30:28 -0700
From: notifications at github.com (Eric Brugger)
This post was imported from the [email protected] email archive

---

The user tried 3.1.2 and it fixed his bug.

---

You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/visit-dav/live-customer-response/issues/9#issuecomment-631762667

URL: <https://elist.ornl.gov/mailman/private/visit-
users/attachments/20200520/64a0e1f1/attachment.html>

---