[Lldb-commits] [PATCH] D103588: [trace] Create a top-level instruction class

wallace added a comment. I've been thinking about what you said and I'm having second thoughts on my implementation. I'll share more context: - I want to work in the short term on reverse debugging and reconstruction of stack traces, for that i'll need to know the instruction type of each inst

[Lldb-commits] [PATCH] D103588: [trace] Create a top-level instruction class

vsk added a comment. In D103588#2806611 , @wallace wrote: >> This approach - of prototyping trace analyses on a vector >> representation and then rewriting them later when scaling issues arise - >> doesn't sound good to me. Even for something simple, li

[Lldb-commits] [PATCH] D103500: [trace][intel-pt] Create basic SB API

wallace updated this revision to Diff 350725. wallace marked 11 inline comments as done. wallace added a comment. Herald added a subscriber: mgorny. address all comments. I appreciate your feedback! - I ended up deleting all the deprecated methods, as I doubt anyone is using them. Intel wrote th

[Lldb-commits] [PATCH] D103588: [trace] Create a top-level instruction class

wallace added a comment. > This approach - of prototyping trace analyses on a vector > representation and then rewriting them later when scaling issues arise - > doesn't sound good to me. Even for something simple, like finding a > backtrace, the window of instructions needed for analysis can e

[Lldb-commits] [PATCH] D103588: [trace] Create a top-level instruction class

vsk added a comment. In D103588#2806512 , @wallace wrote: > Right now we haven't implemented lazy decoding; we are decoding the entire > trace. But that's just because so far we have been working with small traces. > As we progress in this work, we'll s

[Lldb-commits] [lldb] ae1a699 - [LLDB][NFC] Remove parameter names from forward declarations from hand written expressions used in heap.py

Author: Shafik Yaghmour Date: 2021-06-08T14:27:02-07:00 New Revision: ae1a699554cfa01d9fb307a964c3a9f71831a62e URL: https://github.com/llvm/llvm-project/commit/ae1a699554cfa01d9fb307a964c3a9f71831a62e DIFF: https://github.com/llvm/llvm-project/commit/ae1a699554cfa01d9fb307a964c3a9f71831a62e.dif

Re: [Lldb-commits] [lldb] e05b03c - [lldb] Set return status to failed when adding a command error

Hey, David, This commit seems to have caused a new failure in TestRegisters.py, e.g.: http://green.lab.llvm.org/green/blue/organizations/jenkins/lldb-cmake/detail/lldb-cmake/32693/pipeline/ Do you have time to take a look? Jim > On Jun 8, 2021, at 1:41 AM, David Spickett via lldb-commits >

[Lldb-commits] [PATCH] D103588: [trace] Create a top-level instruction class

wallace requested review of this revision. wallace added a comment. > At Apple, we use lldb to navigate instruction traces that contain billions of > instructions. Allocating 16 bytes per instruction simply will not scale for > our workflows. We require the in-memory overhead to be approximately

[Lldb-commits] [PATCH] D103349: [lldb] Don't print script output twice in HandleCommand

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG1a216fb15a18: [lldb] Don't print script output twice in HandleCommand (authored by JDevlieghere). Herald added a project: LLDB. Changed prior to com

[Lldb-commits] [lldb] 1a216fb - [lldb] Don't print script output twice in HandleCommand

Author: Jonas Devlieghere Date: 2021-06-08T13:57:39-07:00 New Revision: 1a216fb15a188f9ac7de6d79267b3cfb2946f792 URL: https://github.com/llvm/llvm-project/commit/1a216fb15a188f9ac7de6d79267b3cfb2946f792 DIFF: https://github.com/llvm/llvm-project/commit/1a216fb15a188f9ac7de6d79267b3cfb2946f792.d

[Lldb-commits] [PATCH] D103652: [lldb][NFC] Refactor name to index maps in Symtab

This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rG64576a1be887: [lldb][NFC] Refactor name to index maps in Symtab (authored by bulbazord). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST A

[Lldb-commits] [lldb] 64576a1 - [lldb][NFC] Refactor name to index maps in Symtab

Author: Alex Langford Date: 2021-06-08T12:36:54-07:00 New Revision: 64576a1be887b7afcacf0534e6c168805fba5677 URL: https://github.com/llvm/llvm-project/commit/64576a1be887b7afcacf0534e6c168805fba5677 DIFF: https://github.com/llvm/llvm-project/commit/64576a1be887b7afcacf0534e6c168805fba5677.diff

[Lldb-commits] [PATCH] D103652: [lldb][NFC] Refactor name to index maps in Symtab

bulbazord updated this revision to Diff 350685. bulbazord added a comment. Rename function :) Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D103652/new/ https://reviews.llvm.org/D103652 Files: lldb/include/lldb/Symbol/Symtab.h lldb/source/Symbo

[Lldb-commits] [PATCH] D103500: [trace][intel-pt] Create basic SB API

mib added a comment. Took a look at the SBAPI changes and left few comments but overall, that part looked good to me. It would be nice if you could delete the depreciated method/class. Comment at: lldb/bindings/interface/SBTrace.i:32-51 + /// deprecated void StopTrace(SBE

[Lldb-commits] [PATCH] D103271: [lldb/Target] Select most relevant frame only in case of signal

jingham added a comment. I don't think it's right to do it in GetStackFrameAtIndex. For instance, suppose lldb stops somewhere, and the user does a backtrace, then they select a different frame on the stack, and then invoke some other command that happens to ask for the frame at index 0. If t

[Lldb-commits] [PATCH] D103271: [lldb/Target] Select most relevant frame only in case of signal

jingham added a comment. "Unwinding" the zero-th frame is really just getting the PC & FP. Since you are very likely to need this information on any given stop, lldb has accelerated versions of the gdb-remote stop reply packets that provide this information up front so this sort of request is

[Lldb-commits] [PATCH] D103588: [trace] Create a top-level instruction class

vsk requested changes to this revision. vsk added a comment. This revision now requires changes to proceed. I'm quite concerned about the design decision to represent a trace as a vector of TraceInstruction. This bakes considerations that appear to be specific to Facebook's usage of IntelPT way

[Lldb-commits] [PATCH] D103500: [trace][intel-pt] Create basic SB API

JDevlieghere added inline comments. Comment at: lldb/bindings/interface/SBTrace.i:32-51 + /// deprecated void StopTrace(SBError &error, lldb::tid_t thread_id); + /// deprecated void GetTraceConfig(SBTraceOptions &options, SBError

[Lldb-commits] [PATCH] D103500: [trace][intel-pt] Create basic SB API

vsk requested changes to this revision. vsk added a subscriber: mib. vsk added a comment. This revision now requires changes to proceed. Thanks, excited to see this work progress! Comment at: lldb/include/lldb/Target/Target.h:1129 /// The trace object. It might be undefine

[Lldb-commits] [PATCH] D103588: [trace] Create a top-level instruction class

wallace updated this revision to Diff 350652. wallace added a comment. I tried to use a ConstString instead of a const char *, but that's not trivial because the compilers wants a safe destructor all the way down, and it seems that this is non-standard across compilers, except for C++17 and onwa

[Lldb-commits] [PATCH] D103500: [trace][intel-pt] Create basic SB API

wallace updated this revision to Diff 350647. wallace added a comment. improved the documentation Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D103500/new/ https://reviews.llvm.org/D103500 Files: lldb/bindings/interface/SBStructuredData.i lldb

[Lldb-commits] [PATCH] D103500: [trace][intel-pt] Create basic SB API

wallace updated this revision to Diff 350641. wallace marked an inline comment as done. wallace added a comment. address all comments Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D103500/new/ https://reviews.llvm.org/D103500 Files: lldb/bindings

[Lldb-commits] [PATCH] D103500: [trace][intel-pt] Create basic SB API

wallace marked 9 inline comments as done. wallace added inline comments. Comment at: lldb/include/lldb/Target/Target.h:1137-1141 + llvm::Expected CreateTrace(); + + /// If a \a Trace object is present, this returns it, otherwise a new Trace is + /// created with \a Trace::Cre

[Lldb-commits] [lldb] c5d56fe - NFC: .clang-tidy: Inherit configs from parents to improve maintainability

Author: David Blaikie Date: 2021-06-08T08:25:59-07:00 New Revision: c5d56fec502f36a0c994835ca23bc93a6c682d95 URL: https://github.com/llvm/llvm-project/commit/c5d56fec502f36a0c994835ca23bc93a6c682d95 DIFF: https://github.com/llvm/llvm-project/commit/c5d56fec502f36a0c994835ca23bc93a6c682d95.diff

[Lldb-commits] [PATCH] D103271: [lldb/Target] Select most relevant frame only in case of signal

PatriosTheGreat updated this revision to Diff 350596. PatriosTheGreat added a comment. Thanks for your suggestions. I tried both of them, however the solution to patch assert recognizer doesn’t solve the original performance issue. Since the performance degradation doesn’t come from the recogniz

[Lldb-commits] [PATCH] D103701: [lldb] Set return status to failed when adding a command error

teemperor added a comment. In D103701#2804840 , @DavidSpickett wrote: > I've gone ahead and landed it, will revert on failures. > > Got patches to remove the reundant calls locally so I'll put those up for > review once this has had time to go through.

[Lldb-commits] [PATCH] D103701: [lldb] Set return status to failed when adding a command error

DavidSpickett added a comment. I've gone ahead and landed it, will revert on failures. Got patches to remove the reundant calls locally so I'll put those up for review once this has had time to go through. (those changes will be fairly mechanical but it's worth someone scanning them for silly m

[Lldb-commits] [PATCH] D103701: [lldb] Set return status to failed when adding a command error

This revision was not accepted when it landed; it landed in state "Needs Review". This revision was landed with ongoing or failed builds. This revision was automatically updated to reflect the committed changes. Closed by commit rGe05b03cf4f45: [lldb] Set return status to failed when adding a com

[Lldb-commits] [lldb] e05b03c - [lldb] Set return status to failed when adding a command error

Author: David Spickett Date: 2021-06-08T09:41:07+01:00 New Revision: e05b03cf4f45ac5ee63c59a3464e7d484884645c URL: https://github.com/llvm/llvm-project/commit/e05b03cf4f45ac5ee63c59a3464e7d484884645c DIFF: https://github.com/llvm/llvm-project/commit/e05b03cf4f45ac5ee63c59a3464e7d484884645c.diff