dtgex2: add 2 to iwork length to prevent OOB interaction in Dtgsy2

[soypat]

Description
This is to fix what seems like a bug, I may be confused however.

Explanation starting from the top/last of the call stack:

• Dtgsy2 requires iwork of length m+n+2

• Dtgsy2 is called from Dtgex2 with arguments m=n1, n=n2

• Dtgex2 requires both n1 and n2 to be either 0, 1, or 2.

• Dtgex2 declares iwork with length ldst=4, so that constrains calling Dtgsy2 with m+n+2<=4

• Dtgexc has a branch (immediately below goto label 20) in which Dtgex2 is called with n1=2 and n2=1

  • This propagates to Dtgsy2 with m=2, n=1, which then fails the iwork length requirement by 1.

  You will notice I added 2 to iwork's length since it is my understanding that both n1 and n2 can both be 2, which would mean a required iwork length of 6 in Dtgsy2.

Checklist

• The documentation has been updated.

No need?

• If the PR solves a specific issue, it is set to be closed on merge.

DNE.

[codecov]

Codecov Report

Patch and project coverage have no change.

Comparison is base (ae2f14f) 0.00% compared to head (0cd886d) 0.00%.

❗ Current head 0cd886d differs from pull request most recent head 710c8bb. Consider uploading reports for the commit 710c8bb to get more accurate results

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #907   +/-   ##
=======================================
  Coverage    0.00%    0.00%           
=======================================
  Files        1918     1918           
  Lines      188614   188614           
=======================================
  Misses     188614   188614           

Files Changed	Coverage Δ
SRC/dtgex2.f	0.00% <ø> (ø)

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[weslleyspereira]

The change makes sense to me. Thanks.

[ilayn]

Probably s flavor also needs this fix?

[soypat]

Probably s flavor also needs this fix?

Done!