search

fermi-ad / controls

Central repo for reporting bugs, making feature requests, managing RFCs, and requesting seminar topics.
https://www-bd.fnal.gov/controls/
2 stars 0 forks source link

Large array data not correct #17

Open beauremus opened 9 months ago

beauremus commented 9 months ago

Describe the bug From @bigsamich:

I think we found a bug somewhere in the stack. We have an array device that is just filling itself with the offset requested. What we have found is that the offset is being jumped back to the 32KB boundary ( What is odd is that this occurs at the 98KB boundary. This should have been sequential.

ACL> read Z:BSTUINT32[0:29999]

24543  24544  24545  24546  24547  24548  24549  24550  24551
24552  24553  24554  24555  24556  24557  8174   8175   8176 
8177   8178   8179   8180   8181   8182   8183   8184   8185 
8186   8187   8188   8189   8190   8191   8192   8193   8194

To Reproduce Steps to reproduce the behavior:

  1. Use ACL with the command read Z:BSTUINT32[0:29999]

Expected behavior This request should report the full array with accurate values. The expected values for the array device are monotonically increasing by 1.

Screenshots If applicable, add screenshots to help explain your problem.

Environment (please complete the following information):

Additional context This came from the #group-central-service Slack channel.

beauremus commented 9 months ago

From @bigsamich:

the plot thickens.....

read Z:BSTUINT32[10000:29999]
rolled over at the 32KB boundary, and then the next offset is random

18176  18177  18178  18179  18180  18181  18182  18183
18184  18185  1802   1803   1804   1805   1806   1807 
1808   1809   1810   1811   1812   1813   1814   1815

this smells like a buffer overflow.....