Skip to content

Conversation

@mikem8361
Copy link
Contributor

@mikem8361 mikem8361 commented Mar 11, 2021

Customer Impact

Incomplete core dumps on Linux are being generated by createdump (via dotnet-dump or unhandled exception env var triggers) causing SOS commands to fail. Multiple customers have reported this problem in 3.1 and 5.0 when their GC heaps have grown larger than 4GB.

Regression?

No. The bug has been in createdump from the beginning.

Risk (see taxonomy)

Low.

Is there a packaging impact?

No.

Full description

The core dump generated for a app that has large GC heaps (>4GB) are don't contain all the memory needed in process. This is because of a 32bit size value overflow; changed to size_t.

Issue: https://devdiv.visualstudio.com/DevDiv/_workitems/edit/1277488?src=WorkItemMention&src-action=artifact_link
Issue: dotnet/diagnostics#1780

@mikem8361 mikem8361 self-assigned this Mar 11, 2021
@ghost
Copy link

ghost commented Mar 11, 2021

Tagging subscribers to this area: @tommcdon
See info in area-owners.md if you want to be subscribed.

Issue Details

The core dump generated for a app that has large GC heaps (>4GB) are don't contain all the memory needed in process. This is because of a 32bit size value overflow; changed to size_t.

Multiple customers have reported this problem in 3.1 and 5.0.

Issue: https://devdiv.visualstudio.com/DevDiv/_workitems/edit/1277488?src=WorkItemMention&src-action=artifact_link and dotnet/diagnostics#1780

Author: mikem8361
Assignees: mikem8361
Labels:

area-Diagnostics-coreclr

Milestone: -

@mikem8361 mikem8361 requested a review from sdmaclea March 11, 2021 22:56
@mikem8361 mikem8361 added the Servicing-consider Issue for next servicing release review label Mar 11, 2021
The core dump generated for a app that has large GC heaps (>4GB) are don't contain all the memory needed in process. This is because of a 32bit size value overflow; changed to size_t.

Multiple customers have reported this problem in 3.1 and 5.0.

Issue: https://devdiv.visualstudio.com/DevDiv/_workitems/edit/1277488?src=WorkItemMention&src-action=artifact_link and dotnet/diagnostics#1780
@jeffschwMSFT jeffschwMSFT removed the Servicing-consider Issue for next servicing release review label Mar 16, 2021
@jeffschwMSFT
Copy link
Member

@mikem8361 can you add the servicing template? We can then take for consideration.

@mikem8361 mikem8361 added the Servicing-consider Issue for next servicing release review label Mar 16, 2021
@mikem8361
Copy link
Contributor Author

@jeffschwMSFT done.

@jeffschwMSFT jeffschwMSFT added this to the 5.0.x milestone Mar 16, 2021
Copy link
Member

@jeffschwMSFT jeffschwMSFT left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Approved. I will take for consideration for 5.0.x

@danmoseley
Copy link
Member

Just curious, do we test this only manually?

@leecow leecow added Servicing-approved Approved for servicing release and removed Servicing-consider Issue for next servicing release review labels Mar 23, 2021
@leecow leecow modified the milestones: 5.0.x, 5.0.6 Mar 23, 2021
@mikem8361
Copy link
Contributor Author

mikem8361 commented Mar 23, 2021 via email

@mikem8361
Copy link
Contributor Author

When should this PR be merged and by whom?

@danmoseley
Copy link
Member

@Anipik merges approved PRs when he opens the branch for the next month. Ideally CI is green

@mikem8361
Copy link
Contributor Author

Thanks Dan. These CI legs keep timing out after multiple retries.

@danmoseley
Copy link
Member

https://dev.azure.com/dnceng/public/_build/results?buildId=1040713&view=logs&jobId=c4eeb0cc-e033-5610-ec93-179a8329d6d1&j=97136b44-3c84-594d-5381-1e1cec29c916&t=fb40d87e-117e-573c-3759-a77950be96e9

2021-04-02T03:35:46.1443375Z Total Files: 1, Processed: 1, Skipped: 0, Failed: 0, Download time: 220.773 secs, Download size: 216.088MB
2021-04-02T05:30:19.5998117Z ##[error]The operation was canceled.
2021-04-02T05:30:19.6012343Z ##[section]Finishing: Download artifacts for all platforms

@mthalman any idea why the installer legs time out here?

@mthalman
Copy link
Member

mthalman commented Apr 5, 2021

https://dev.azure.com/dnceng/public/_build/results?buildId=1040713&view=logs&jobId=c4eeb0cc-e033-5610-ec93-179a8329d6d1&j=97136b44-3c84-594d-5381-1e1cec29c916&t=fb40d87e-117e-573c-3759-a77950be96e9

2021-04-02T03:35:46.1443375Z Total Files: 1, Processed: 1, Skipped: 0, Failed: 0, Download time: 220.773 secs, Download size: 216.088MB
2021-04-02T05:30:19.5998117Z ##[error]The operation was canceled.
2021-04-02T05:30:19.6012343Z ##[section]Finishing: Download artifacts for all platforms

@mthalman any idea why the installer legs time out here?

I assume you mean @mmitche?

@danmoseley
Copy link
Member

@mthalman for some reason I thought you owned the installer area of this repo, but I see it is @NikolaMilosavljevic

@NikolaMilosavljevic
Copy link
Member

@mthalman for some reason I thought you owned the installer area of this repo, but I see it is @NikolaMilosavljevic

The log is inconclusive. @mmitche are you aware of timing issues with this step?

2021-04-02T03:32:04.2199208Z Task         : Download build artifacts

https://dev.azure.com/dnceng/9ee6d478-d288-47f7-aacc-f6e6d082ae6d/_apis/build/builds/1040713/logs/2057

@Anipik Anipik merged commit 4bd646f into dotnet:release/5.0 Apr 6, 2021
@mikem8361 mikem8361 deleted the createdump50 branch April 10, 2021 18:17
@ghost ghost locked as resolved and limited conversation to collaborators May 10, 2021
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

8 participants