Shulou(Shulou.com)06/01 Report--

This article mainly analyzes the relevant knowledge points of Bug instance analysis in the RyuJIT compiler of NET 4.6. the content is detailed and easy to understand, and the operation details are reasonable, which has a certain reference value. If you are interested, you might as well follow the editor and learn more about "Bug instance Analysis in the RyuJIT Compiler of .NET 4.6".

Two more serious bug have been found in the .NET 4.6 RyuJIT compiler, and developer Andrey Akinshin has documented the first bug and provided proof-of-concept code in C # to explain the problems. It was a StackOverflow user "BrandonAGr" who left a message on the site that inspired Akinshin to write this code.

Akinshin uses a nullable Boolean type in the example, which shows how RyuJIT causes an error in the generated assembly when it tries to generate optimized code. Judging from the results of this example, it produces an error the first time it is run, but everything is fine the second time, which is even more troubling. In conclusion, Akinshin said: "RyuJIT seems to have a rather tricky optimization bug. The generation of this bug depends on a large number of additional conditions, and every change to the source code destroys its reproducibility."

Fortunately, schellap, a developer from Microsoft, has submitted a fix for this bug, but as with the problem reported earlier this week, no binary patch has been released yet. Users expressed their frustrations and worries about the current situation of RyuJIT. According to user "EasmoneNerbonne", based on the reported bug, "… it seems that the best thing to do is to completely stop using [RyuJIT] in the near future, rather than just disabling the tail call feature." Akinshin agrees. User Daniel Marbach also said: "… the blog about RyuJIT should update these bug, and Microsoft should issue a public statement clearly: do not use RyuJIT in a production environment." (the original text is emphasized in uppercase)

The second bug is reported by GitHub user Hyunho Richard Lee, who shows the bug with an example of F# code: a program that should be running normally has an error (a FatalExecutionEngineError exception is thrown). According to Lee, for F# 3.1 programs that work properly and target .NET 4.5.1 (or 4.5,4.5.2), there will be problems after installing .NET 4.6 RC.

Lee summed up this problem for F# developers:

If F# developers execute the Seq.distinct method on a KeyValuePair-type Seq in any part of the code of an existing application, they must control the runtime environment to ensure that .NET 4.6 is not installed in the deployment environment. Another approach is to immediately reset the goals of the project and then immediately redeploy the projects.

Microsoft developer sivarv hinted that the problem had been fixed in a submission, but as mentioned above, the release date of any binary patch pack has not been publicly announced.

This is the end of the introduction of "Bug instance Analysis in the RyuJIT Compiler of .NET 4.6". More related content can be searched for previous articles, hoping to help you answer questions and questions, please support the website!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.