Replace CodeManState with cache in EECodeInfo for x86. by filipnavara · Pull Request #114170 · dotnet/runtime

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

filipnavara · 2025-04-02T20:25:10Z

On all other platforms CodeManState was unused and taking space in StackFrameIterator.

Optimize some code paths to cache GC info early or decode the method size without decoding the whole table.

dotnet-policy-service · 2025-04-02T20:25:50Z

Tagging subscribers to this area: @mangod9
See info in area-owners.md if you want to be subscribed.

filipnavara · 2025-04-02T20:30:17Z

This is not an universal win.

It makes StackFrameIterator on x86 slightly larger because it contains two EECodeInfo classes. It may also increase stack usage in places where EECodeInfo is used but most of those seem to access the GC info at some point anyway so embedding the buffer for the cache doesn't make things worse.

On other platforms it saves an unused space. There seem to be no downsides.

It gets rid of one x86 quirk and makes it easier to reuse the cache in FEATURE_EH_FUNCLETS builds, currently to very small benefit (about 5% speed up of stack iterators).

Thoughts?

(I am aware of the build issues due to mismatching offsets; not worth resolving if this whole direction gets abandoned)

jkotas · 2025-04-03T17:55:01Z

LGTM! I like that it gets rid of the unused cache on !x86.

filipnavara · 2025-04-03T19:09:21Z

LGTM! I like that it gets rid of the unused cache on !x86.

Thanks for having a look. I made some progress on this locally and it helped me improve the performance on x86+funclet platforms. I'll clean up the PR soon.

On all other platforms CodeManState was unused and taking space in StackFrameIterator. Optimize some code paths to cache GC info early or decode the method size without decoding the whole table. Move PCTAddr (Eip pointer) into KNONVOLATILE_CONTEXT_POINTERS_EX to avoid recomputing it when switching between callee and caller contexts.

…eep the helpers

filipnavara · 2025-04-08T14:40:27Z

+inline TADDR GetRegdisplayPCTAddr(REGDISPLAY *display)
+{
+ return display->PCTAddr;
+}
+
 inline void SetRegdisplayPCTAddr(REGDISPLAY *display, TADDR addr)


I initially tried to optimize how the PCTAddr is stored for callee context to avoid recomputation. Unfortunately it results in some bugs (eg. CopyRegDisplay doesn't transfer the information correctly, IsInCalleesFrames depends on quirks in the initial state, etc.) and I opted not to pursue this optimization at this time. I left the helper methods in place to make it easier to address this in future and to create a central place where we can _ASSERTE on invalid values.

…e and thus avoid repeated GetGCInfoToken() calls

jkotas

LGTM. Thank you!

@janvorli Could you please take a look as well?

janvorli

LGTM, thank you!

janvorli · 2025-04-09T11:48:51Z

/ba-g the test failure has been happening consistently on all runs since yesterday. #114426

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

ghost added the area-VM-coreclr label Apr 2, 2025

dotnet-policy-serviceBot added the community-contribution Indicates that the PR has been added by a community member label Apr 2, 2025

filipnavara requested a review from jkotas April 2, 2025 20:30

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/threads.h Outdated

filipnavara commented Apr 2, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp Outdated

filipnavara force-pushed the no-codemanstate branch from fbe883d to 16b67b1Compare April 8, 2025 08:15

filipnavara changed the title ~~RFC: Replace CodeManState with cache in EECodeInfo for x86.~~Replace CodeManState with cache in EECodeInfo for x86.Apr 8, 2025

filipnavara added 2 commits April 8, 2025 16:13

Revert the PCTAddr move into KNONVOLATILE_CONTEXT_POINTERS->Eip but k…
363f5f7
…eep the helpers

filipnavara force-pushed the no-codemanstate branch from ca82fc0 to 363f5f7Compare April 8, 2025 14:13

filipnavara marked this pull request as ready for review April 8, 2025 14:16

filipnavara commented Apr 8, 2025
View reviewed changes

filipnavara mentioned this pull request Apr 8, 2025
New exception handling on win-x86 #113985
Closed
6 tasks

Improve the GC info cache by caching the table pointer instead of siz…
1aaf657
…e and thus avoid repeated GetGCInfoToken() calls

jkotas approved these changes Apr 9, 2025
View reviewed changes

janvorli reviewed Apr 9, 2025
View reviewed changes

Comment threadsrc/coreclr/vm/stackwalk.cpp
Comment threadsrc/coreclr/vm/stackwalk.h

janvorli approved these changes Apr 9, 2025
View reviewed changes

janvorli merged commit 2b922d0 into dotnet:mainApr 9, 2025

filipnavara deleted the no-codemanstate branch April 9, 2025 11:55

tommcdon mentioned this pull request May 8, 2025
Fix incorrect codemanager implementation on x86 #115391
Merged

github-actionsBot locked and limited conversation to collaborators May 10, 2025

Uh oh!

Conversation

filipnavara commented Apr 2, 2025• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dotnet-policy-serviceBot commented Apr 2, 2025

Uh oh!

filipnavara commented Apr 2, 2025• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

jkotas commented Apr 3, 2025

Uh oh!

filipnavara commented Apr 3, 2025

Uh oh!

filipnavaraApr 8, 2025

Choose a reason for hiding this comment

Uh oh!

jkotas left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

janvorli left a comment

Choose a reason for hiding this comment

Uh oh!

janvorli commented Apr 9, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

filipnavara commented Apr 2, 2025• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dotnet-policy-serviceBot commented Apr 2, 2025

Uh oh!

filipnavara commented Apr 2, 2025• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

jkotas commented Apr 3, 2025

Uh oh!

filipnavara commented Apr 3, 2025

Uh oh!

filipnavaraApr 8, 2025

Choose a reason for hiding this comment

Uh oh!

jkotas left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

janvorli left a comment

Choose a reason for hiding this comment

Uh oh!

janvorli commented Apr 9, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

filipnavara commented Apr 2, 2025• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

filipnavara commented Apr 2, 2025•
edited
Loading

filipnavara commented Apr 2, 2025•
edited
Loading

filipnavara commented Apr 2, 2025•
edited
Loading

filipnavara commented Apr 2, 2025•
edited
Loading

filipnavara commented Apr 2, 2025•
edited
Loading

filipnavara commented Apr 2, 2025•
edited
Loading

filipnavara commented Apr 2, 2025•
edited
Loading

filipnavara commented Apr 2, 2025•
edited
Loading