- 21 6月, 2017 2 次提交
-
-
由 Heejae Chang 提交于
-
由 Heejae Chang 提交于
-
- 20 6月, 2017 3 次提交
-
-
由 Heejae Chang 提交于
-
由 Heejae Chang 提交于
-
由 Heejae Chang 提交于
-
- 17 6月, 2017 2 次提交
-
-
由 Heejae Chang 提交于
-
由 Heejae Chang 提交于
made diagnostic engine to verify span given to diagnostic from analyzers. and added explicit check in tagger for out of range
-
- 16 6月, 2017 5 次提交
-
-
由 Ravi Chande 提交于
Fix crash in VB type inferrer
-
由 David Poeschl 提交于
Accessibility for Inline Rename in the Editor
-
由 Ivan Basov 提交于
This reverts commit 2fba37a6.
-
由 Andy Gocke 提交于
By design, when data flow analyzes local functions it does so devoid of surrounding context and records possibly unassigned variables rather than reporting a diagnostic. Thus, all captured variables are expected to be marked unassigned during data flow analysis without reporting a diagnostic. However, reporting a diagnostic is not the only side effect of running data flow analysis. By calling virtual methods data flow analysis informs derived types of unassigned variables. While the diagnostic is suppressed and recorded, currently the virtual method is still called. This has negative consequences for region analysis, which considers these calls as indications of variables flowing out of the given region, which may not be the case for local functions. This PR changes data flow analysis to only call the ReportUnassigned virtual method if not inside a local function, excluding local functions from all unassignment reporting. Fixes #17165, #18347
-
由 David Poeschl 提交于
Fixes https://devdiv.visualstudio.com/DevDiv/_workitems?id=391624&_a=edit - Inline Rename fields now consist of a themeable/customizable background, foreground, and border. The border is used in High Contrast. - Resolved and Unresolved conflict borders can now be themed & customized. - Resolved and Unresolved conflict borders now have different dash styles so they are distinct from each other and regular Inline Rename fields. The Dashboard shows a key indicating which type of conflict corresponds to which border dash style.
-
- 15 6月, 2017 4 次提交
-
-
由 Ravi Chande 提交于
Skip repl VSI tests
-
由 Jason Malinowski 提交于
Log telemetry if we see ITextBufferEdit.Apply() throwing exceptions
-
由 Ravi Chande 提交于
Failure tracked by https://github.com/dotnet/roslyn/issues/20219
-
由 Jared Parsons 提交于
Looks like there is still fall out from the licensing issue. VS integration tests still failing at 100% due to a timeout issue. Change them to not run by default while we look into this. Can still be run manually for testing with test vsintegration please
-
- 14 6月, 2017 3 次提交
-
-
由 Ivan Basov 提交于
-
由 Jared Parsons 提交于
Move language settings after SDK import
-
由 Jared Parsons 提交于
The new SDK unconditionally sets a number of MSBulid properties including many that we depend on in Roslyn. This is a break from the old MSBuild behavior which respected initial values. https://github.com/dotnet/sdk/issues/752 The behavior in 2.0 is changing to go back to the old model of respecting initial values. Roslyn though continues to often use the SDK installed with Visual Studio. That can be a variety of values from 2.0 prebuilds, to various drops of 1.0. As a result the success / failure of our build can depend a lot on which version of VS is installed and by consequence which new SDK comes with it. Subtle differences in the SDK can change the values we end up compiling with in our project. We've been guarding against this by directly re-setting values we knew to be overidden in some version of the SDK. Now though I'm worried this isn't an adequete solution for the following reasons. - Just this week I got new warnings in the build by using a new VS - The Microbuild lab is aggressively grabbing new builds of VS which means the VS SDK is also potentially changing. After some thought I decided to simply move all of our langauge specific settings after we import the core targets. Given that we unconditionally set the values it becomes irrelevant what the SDK does or does not respect.
-
- 13 6月, 2017 5 次提交
-
-
由 CyrusNajmabadi 提交于
Be resilient to running with System.IEquatable can not be found. Fixes #20158
-
由 CyrusNajmabadi 提交于
Fix race with boosting OOP priority. Also, ensure FindRef calls are boosted.
-
由 CyrusNajmabadi 提交于
-
由 CyrusNajmabadi 提交于
-
由 CyrusNajmabadi 提交于
-
- 10 6月, 2017 10 次提交
-
-
由 Kevin Pilch 提交于
Force load projects that failed design time builds, not ones that opt out of LSL
-
由 Manish Vasani 提交于
Remove strong reference to AnalyzerOptions in the analyzer manager ex…
-
由 Jared Parsons 提交于
Track down NuGet cache corruption
-
由 Brett V. Forsgren 提交于
Revert "allow waiters to be consumed from internal VS tests"
-
由 Tomáš Matoušek 提交于
Refactor tests
-
由 Manish Vasani 提交于
Add a unit test to ensure that analyzer execution does not leak solution instances from preview workspace.
-
由 Jared Parsons 提交于
This change should help us pin down why we are seeing NuGet cache corruption issues in our Jenkins job. It adds a check to three places in our script: - Start - Post restore - End Depending on when this check fails it should give us a better picture of where the corruption is occuring. https://github.com/dotnet/roslyn/issues/19882
-
由 Brett V. Forsgren 提交于
-
由 Jason Malinowski 提交于
Hack around PublicAPI item group support in the new project system
-
由 Manish Vasani 提交于
Move Roslyn assemblies to consume newer OptimizationData
-
- 09 6月, 2017 3 次提交
-
-
由 Jason Malinowski 提交于
This is a workaround for dotnet/project-system#2160 where AdditionalFileItemNames isn't quite behaving correctly with the IDE. This can be reverted once that bug is fixed.
-
由 Julien Couvreur 提交于
-
由 Andy Gocke 提交于
This reverts commit 8a1ca479 and moves publish earlier in the build pipeline (before signing) to prevent published binaries from overwriting already built and signed binaries.
-
- 08 6月, 2017 3 次提交
-
-
由 Manish Vasani 提交于
Fixes #20065
-
由 Manish Vasani 提交于
This should likely fix an RPS regression (VSO 440011) related to additional jitting of Roslyn assemblies.
-
由 Kevin Pilch 提交于
The solution build manager will handle the opted out ones, but not the failed design time build ones. Fixes #19407.
-