Re: [gem5-dev] gem5 tags rework

> In the case that nothing matches, you can always use commit --no-verify to > skip that verification step. I am not a fan of this solution. If it is not something entirely new, it must match. Otherwise, a new tag should be proposed, along with the sequence of patches that use it. > Maybe

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Implement reading vector registers with readVecReg.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22117 ) Change subject: fastmodel: Implement reading vector registers with readVecReg. .. fastmodel: Implement reading

[gem5-dev] Change in gem5/gem5[master]: arm: Implement inst count events in the IRIS thread contexts.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22115 ) Change subject: arm: Implement inst count events in the IRIS thread contexts. .. arm: Implement inst count even

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Add an address translation mechanism to the ThreadContext.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22118 ) Change subject: fastmodel: Add an address translation mechanism to the ThreadContext. .. fastmodel: Add an addr

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Implement port proxies.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22120 ) Change subject: fastmodel: Implement port proxies. .. fastmodel: Implement port proxies. This plumbing is simple

[gem5-dev] Change in gem5/gem5[master]: cpu: Get rid of the nextInstEventCount method.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22114 ) Change subject: cpu: Get rid of the nextInstEventCount method. .. cpu: Get rid of the nextInstEventCount method.

[gem5-dev] Change in gem5/gem5[master]: cpu: Delegate comInstEventQueue methods to the ThreadContexts.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22108 ) Change subject: cpu: Delegate comInstEventQueue methods to the ThreadContexts. .. cpu: Delegate comInstEventQue

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Implement PC based events.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22122 ) Change subject: fastmodel: Implement PC based events. .. fastmodel: Implement PC based events. These use the IRIS

[gem5-dev] Change in gem5/gem5[master]: cpu: Create a PCEventScope class to abstract the scope of PCEvents.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22099 ) Change subject: cpu: Create a PCEventScope class to abstract the scope of PCEvents. .. cpu: Create a PCEventSco

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Create a TLB model which uses IRIS to do translations.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22119 ) Change subject: fastmodel: Create a TLB model which uses IRIS to do translations. .. fastmodel: Create a TLB mo

[gem5-dev] Change in gem5/gem5[master]: cpu: Get rid of the serviceInstCountEvents method.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22113 ) Change subject: cpu: Get rid of the serviceInstCountEvents method. .. cpu: Get rid of the serviceInstCountEvents m

[gem5-dev] Change in gem5/gem5[master]: cpu: Access inst events through ThreadContext instead of the CPU.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22109 ) Change subject: cpu: Access inst events through ThreadContext instead of the CPU. .. cpu: Access inst events th

[gem5-dev] Change in gem5/gem5[master]: sim: Make the System object a PCEventScope.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22101 ) Change subject: sim: Make the System object a PCEventScope. .. sim: Make the System object a PCEventScope. This a

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Plumb the ITB and DTB through the IRIS thread context.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22116 ) Change subject: fastmodel: Plumb the ITB and DTB through the IRIS thread context. .. fastmodel: Plumb the ITB a

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Implement getCurrentInstCount.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22111 ) Change subject: fastmodel: Implement getCurrentInstCount. .. fastmodel: Implement getCurrentInstCount. This uses

[gem5-dev] Change in gem5/gem5[master]: cpu: Switch off of the CPU's comInstEventQueue.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22110 ) Change subject: cpu: Switch off of the CPU's comInstEventQueue. .. cpu: Switch off of the CPU's comInstEventQueue.

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Use getCurrentInstCount for totalInsts().

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22112 ) Change subject: fastmodel: Use getCurrentInstCount for totalInsts(). .. fastmodel: Use getCurrentInstCount for tot

[gem5-dev] Change in gem5/gem5[master]: fastmodel: Tell fast model not to shutdown when time stops.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22121 ) Change subject: fastmodel: Tell fast model not to shutdown when time stops. .. fastmodel: Tell fast model not to s

[gem5-dev] Change in gem5/gem5[master]: cpu: Make accesses to comInstEventQueue indirect through methods.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22107 ) Change subject: cpu: Make accesses to comInstEventQueue indirect through methods. .. cpu: Make accesses to comI

[gem5-dev] Change in gem5/gem5[master]: cpu, sim: Delegate PCEvent scheduling from Systems to ThreadContexts.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22106 ) Change subject: cpu,sim: Delegate PCEvent scheduling from Systems to ThreadContexts. .. cpu,sim: Delegate PCEve

[gem5-dev] Change in gem5/gem5[master]: cpu: Stop checking for PC changes when servicing a PCEventQueue.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22100 ) Change subject: cpu: Stop checking for PC changes when servicing a PCEventQueue. .. cpu: Stop checking for PC c

[gem5-dev] Change in gem5/gem5[master]: cpu: Pass the address to check into the PCEventQueue service method.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22102 ) Change subject: cpu: Pass the address to check into the PCEventQueue service method. .. cpu: Pass the address t

[gem5-dev] Change in gem5/gem5[master]: cpu: Make the ThreadContext a PCEventScope.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22105 ) Change subject: cpu: Make the ThreadContext a PCEventScope. .. cpu: Make the ThreadContext a PCEventScope. Both t

[gem5-dev] Change in gem5/gem5[master]: cpu, sim: Get rid of a bunch of conditional compilation for PCEvents.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22104 ) Change subject: cpu,sim: Get rid of a bunch of conditional compilation for PCEvents. .. cpu,sim: Get rid of a b

[gem5-dev] Change in gem5/gem5[master]: cpu: Don't print the CPU name when a (Break|Panic)PCEvent happens.

Gabe Black has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22103 ) Change subject: cpu: Don't print the CPU name when a (Break|Panic)PCEvent happens. .. cpu: Don't print the CPU

Re: [gem5-dev] gem5 tags rework

Yeah, if it's on the client side and you can --no-verify, that more or less puts it in the warning category. The problem with --no-verify is that it's an all or nothing thing, so if I know (for instance) that I'll be systematically breaking some commit rule (importing/changing externally sourced co

Re: [gem5-dev] gem5 tags rework

We should add systemc as a tag and set Gabe as the maintainer :). I like the idea of the default to be a strict set of tags. This would force the contributor to think about who should be tagged in the review as well (it would be great if we could automate this!). In the case that nothing matches,

Re: [gem5-dev] gem5 tags rework

I'm not sure if there's a lot of value in having a very strictly set defined set of tags. That's sort of (but not entirely) like having a fixed set of email subject lines. Sure, it makes it easier to group things, but not everything will fit in the predefined categories. I think needless difference

[gem5-dev] Change in gem5/gem5[master]: dev: Adding support for large BAR

Pouya Fotouhi has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22082 ) Change subject: dev: Adding support for large BAR .. dev: Adding support for large BAR Checks if a BAR is used

Re: [gem5-dev] gem5 tags rework

It seems like there may be some value in adding "systemc" as an official tag, but, from what I can see, most of the other incorrect tags are mapped pretty cleanly to existing ones. Would anyone have any objection to adding "systemc" as a tag in MAINTAINERS? In regards to server-side hooks: it s

Re: [gem5-dev] Elf loading problem

*With different access permissions From: gem5-dev on behalf of Giacomo Travaglini Sent: 22 October 2019 17:41 To: Gabe Black ; gem5 Developer List Subject: Re: [gem5-dev] Elf loading problem Sorry Gabe, forgot to address your other question. We have an interna

Re: [gem5-dev] Elf loading problem

Sorry Gabe, forgot to address your other question. We have an internal set of "unusual" binaries, which are testing architectural features. Some of these tests mark pages with different address access probably to check accesses which are crossing page boundaries, so they are built using 1 segme

[gem5-dev] Change in gem5/gem5[master]: configs: Clean setupBootLoader signature

Giacomo Travaglini has submitted this change. ( https://gem5-review.googlesource.com/c/public/gem5/+/22002 ) Change subject: configs: Clean setupBootLoader signature .. configs: Clean setupBootLoader signature This is because

[gem5-dev] Change in gem5/gem5[master]: sim-se: port DPRINTF_SYSCALL to use the automatic DPRINTF flag

Ciro Santilli has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22008 ) Change subject: sim-se: port DPRINTF_SYSCALL to use the automatic DPRINTF flag .. sim-se: port DPRINTF_SYSCA

[gem5-dev] Change in gem5/gem5[master]: base: create DPRINTF_UNCONDITIONAL

Ciro Santilli has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22005 ) Change subject: base: create DPRINTF_UNCONDITIONAL .. base: create DPRINTF_UNCONDITIONAL This is similar to DP

[gem5-dev] Change in gem5/gem5[master]: cpu: prefix ExecEnable to the native trace to match DPRINTF

Ciro Santilli has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22007 ) Change subject: cpu: prefix ExecEnable to the native trace to match DPRINTF .. cpu: prefix ExecEnable to the na

[gem5-dev] Change in gem5/gem5[master]: base: generalize ExecTicks to all messages with FmtTicksOff

Ciro Santilli has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22006 ) Change subject: base: generalize ExecTicks to all messages with FmtTicksOff .. base: generalize ExecTicks to al

[gem5-dev] Change in gem5/gem5[master]: base: add the --debug-flag to DPRINTF output

Ciro Santilli has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22004 ) Change subject: base: add the --debug-flag to DPRINTF output .. base: add the --debug-flag to DPRINTF output T

[gem5-dev] Change in gem5/gem5[master]: sim: prefix --debug-flags Event logs with the flag name

Ciro Santilli has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22009 ) Change subject: sim: prefix --debug-flags Event logs with the flag name .. sim: prefix --debug-flags Event logs

[gem5-dev] Change in gem5/gem5[master]: base: add the FmtStackTrace debug option

Ciro Santilli has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22003 ) Change subject: base: add the FmtStackTrace debug option .. base: add the FmtStackTrace debug option If given,

[gem5-dev] Change in gem5/gem5[master]: configs: Do not assume bootmem is a System child

Giacomo Travaglini has submitted this change. ( https://gem5-review.googlesource.com/c/public/gem5/+/22001 ) Change subject: configs: Do not assume bootmem is a System child .. configs: Do not assume bootmem is a System child

[gem5-dev] Change in gem5/gem5[master]: dev-arm, configs: Using _on_chip_memory for on chip memory

Giacomo Travaglini has submitted this change. ( https://gem5-review.googlesource.com/c/public/gem5/+/22000 ) Change subject: dev-arm, configs: Using _on_chip_memory for on chip memory .. dev-arm, configs: Using _on_chip_memory

Re: [gem5-dev] Elf loading problem

Right I will push a new version where the segments are named after the index. This is in any case how they are called when you parse the elf binary with the readelf --segments. If someone wants to know which sections a segment contains, could just cross reference the segment name in gem5 with th

[gem5-dev] Change in gem5/gem5[master]: [WIP] fastmodel: Couple of fixes to bring up CPU1

Chun-Chen TK Hsu has uploaded this change for review. ( https://gem5-review.googlesource.com/c/public/gem5/+/22059 ) Change subject: [WIP] fastmodel: Couple of fixes to bring up CPU1 .. [WIP] fastmodel: Couple of fixes to bri