Uh oh!
There was an error while loading. Please reload this page.
Add arm64 support for EnC - #69679
Conversation
This adds support for EnC on arm64. A couple of notes on the implementation compared to x64: - On x64 we get the fixed stack size from unwind info. However, for the frames we set up on arm64 for EnC it is not possible to extract the frame size from there because their prologs generally look like stp fp, lr, [sp,#-16]! mov fp, sp sub sp, sp, dotnet#144 with unwind codes like the following: set_fp; mov fp, sp save_fplr_x #1 (0x01); tp fp, lr, [sp, #-16]! As can be seen, it is not possible to get the fixed stack size from unwind info in this case. Instead we pass it through the GC info that already has a section for EnC data. - On arm64 the JIT is required to place the PSPSym at the same offset from caller-SP for both the main function and for funclets. Due to this we try to allocate the PSPSym as early as possible in the main function and we must take some care in funclets. However, this conflicts with the EnC frame header that the JIT uses to place values that must be preserved on EnC transitions. This is currently callee-saved registers and the MonitorAcquired boolean. Before this change we were allocating PSPSym above (before) the monitor acquired boolean, but we now have to allocate MonitorAcquired first, particularly because the size of the preserved header cannot change on EnC transitions, while the PSPSym can disappear or appear. This changes frame allocation slightly for synchronized functions.
ghost
commented
May 23, 2022
Tagging subscribers to this area: @JulieLeeMSFT Issue DetailsThis adds support for EnC on arm64. A couple of notes on the
|
| CORDbgSetInstruction((CORDB_ADDRESS_TYPE *)patchBypass, 0xd503201f); //Add Nop in buffer | ||
| SetReg(context, RegNum, finalAddr); | ||
| return patchBypass; | ||
| } |
There was a problem hiding this comment.
I am including a bugfix to the stepper. On arm64 we would not handle adr/adrp correctly which causes issues when running diagnostics tests.
AndyAyersMS
left a comment
There was a problem hiding this comment.
Do we need to update the R2R version because of the gc info change? I suppose we don't as this new format is only there with ENC, which we won't see in R2R.
Uh oh!
There was an error while loading. Please reload this page.
jakobbotsch
commented
May 23, 2022
Right, I checked with @davidwrighton and he was under the same belief. |
jakobbotsch
commented
May 23, 2022
Looks like there is an OSR + synchronized assert I'll need to sort out. |
OSR methods reuse the bool created by the tier 0 frame so we should not allocate space for it.
This adds support for EnC on arm64. A couple of notes on the
implementation compared to x64:
On x64 we get the fixed stack size from unwind info. However, for the
frames we set up on arm64 for EnC it is not possible to extract the
frame size from there because their prologs generally look like
stp fp, lr, [sp,#-16]!
mov fp, sp
sub sp, sp, # 144
with unwind codes like the following:
set_fp; mov fp, sp
save_fplr_x # 1 (0x01); stp fp, lr, [sp, #-16]!
As can be seen, it is not possible to get the fixed stack size from
unwind info in this case. Instead we pass it through the GC info that
already has a section for EnC data.
On arm64 the JIT is required to place the PSPSym at the same offset
from caller-SP for both the main function and for funclets. Due to
this we try to allocate the PSPSym as early as possible in the main
function and we must take some care in funclets. However, this
conflicts with the EnC frame header that the JIT uses to place values
that must be preserved on EnC transitions. This is currently
callee-saved registers and the MonitorAcquired boolean.
Before this change we were allocating PSPSym above (before) the
monitor acquired boolean, but we now have to allocate MonitorAcquired
first, particularly because the size of the preserved header cannot
change on EnC transitions, while the PSPSym can disappear or appear.
This changes frame allocation slightly for synchronized functions.