Roslyn troubleshooting

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Roslyn troubleshooting

Neale Ferguson
+ADw-html+AD4- +ADw-head+AD4- +ADw-/head+AD4- +ADw-body style+AD0AIg-word-wrap: break-word+ADs- -webkit-nbsp-mode: space+ADs- -webkit-line-break: after-white-space+ADs- font-family: Calibri, sans-serif+ADs- font-size: 14px+ADs- color: rgb(0, 0, 0)+ADsAIgA+- +ADw-div+AD4-The drop of the Roslyn 2.3.0 beta2 update stopped the +IBw-Little Endian Only+IB0- message appearing when using it on s390x (big endian). However, after this message is displayed:+ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4- +ADw-div+AD4-Microsoft (R) Visual C+ACM- Compiler version 2.3.0.61801 (3722bb71)+ADw-/div+AD4- +ADw-div+AD4-Copyright (C) Microsoft Corporation. All rights reserved.+ADw-/div+AD4- +ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4-The process just waits forever. The worker thread(s) appear to have started but found nothing and the main thread keeps on waiting.+ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4-Looking at the traces from x86+AF8-64 and s390x when invoking csc I see that the paths diverge when FindNewWork() returns false on s390x but true on the Intel. What is the best strategy for trying to determine what+IBk-s going on? I assume I could add Console.WriteLine() to the System.Threading.Task code to see if things are being queued properly, but are there better options?+ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4-Neale+ADw-/div+AD4- +ADw-/body+AD4- +ADw-/html+AD4-
_______________________________________________
Mono-devel-list mailing list
[hidden email]
http://lists.dot.net/mailman/listinfo/mono-devel-list
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: Roslyn troubleshooting

Neale Ferguson
+ADw-html+AD4- +ADw-head+AD4- +ADw-/head+AD4- +ADw-body style+AD0AIg-word-wrap: break-word+ADs- -webkit-nbsp-mode: space+ADs- -webkit-line-break: after-white-space+ADs- color: rgb(0, 0, 0)+ADs- font-size: 14px+ADs- font-family: Calibri, sans-serif+ADsAIgA+- +ADw-div+AD4-I was able to find my problem +IBM- I had inadvertently changed the OP+AF8-ATOMIC+AF8-Ix implementation. Once fixed I am back to the Little Endian issue, which is fine of the time being.+ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-span id+AD0AIg-OLK+AF8-SRC+AF8-BODY+AF8-SECTION+ACIAPg- +ADw-div style+AD0AIg-font-family:Calibri+ADs- font-size:11pt+ADs- text-align:left+ADs- color:black+ADs- BORDER-BOTTOM: medium none+ADs- BORDER-LEFT: medium none+ADs- PADDING-BOTTOM: 0in+ADs- PADDING-LEFT: 0in+ADs- PADDING-RIGHT: 0in+ADs- BORDER-TOP: +ACM-b5c4df 1pt solid+ADs- BORDER-RIGHT: medium none+ADs- PADDING-TOP: 3pt+ACIAPg- +ADw-br+AD4- +ADw-/div+AD4- +ADw-div+AD4- +ADw-div style+AD0AIg-word-wrap: break-word+ADs- -webkit-nbsp-mode: space+ADs- -webkit-line-break: after-white-space+ADs- font-family: Calibri, sans-serif+ADs- font-size: 14px+ADs- color: rgb(0, 0, 0)+ADsAIgA+- +ADw-div+AD4-The drop of the Roslyn 2.3.0 beta2 update stopped the +IBw-Little Endian Only+IB0- message appearing when using it on s390x (big endian). However, after this message is displayed:+ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4- +ADw-div+AD4-Microsoft (R) Visual C+ACM- Compiler version 2.3.0.61801 (3722bb71)+ADw-/div+AD4- +ADw-div+AD4-Copyright (C) Microsoft Corporation. All rights reserved.+ADw-/div+AD4- +ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4-The process just waits forever. The worker thread(s) appear to have started but found nothing and the main thread keeps on waiting.+ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4-Looking at the traces from x86+AF8-64 and s390x when invoking csc I see that the paths diverge when FindNewWork() returns false on s390x but true on the Intel. What is the best strategy for trying to determine what+IBk-s going on? I assume I could add Console.WriteLine() to the System.Threading.Task code to see if things are being queued properly, but are there better options?+ADw-/div+AD4- +ADw-div+AD4APA-br+AD4- +ADw-/div+AD4- +ADw-div+AD4-Neale+ADw-/div+AD4- +ADw-/div+AD4- +ADw-/div+AD4- +ADw-/span+AD4- +ADw-/body+AD4- +ADw-/html+AD4-
_______________________________________________
Mono-devel-list mailing list
[hidden email]
http://lists.dot.net/mailman/listinfo/mono-devel-list
Loading...