Set register dep to exclude GPR0 for temp2Reg in inlineCompareAndSet

[rmnattas]

The temp2Reg is used by VMnonNullSrcWrtBarCardCheckEvaluator in some cases in a stbx instruction, which can't have GPR0 as the RA argument or it will act upon it as a null value.
This can cause an unexpected register spill in the middle of an internal control-flow sequence that might not be fully executed [1].

The inserted exclusion is sufficiently conditioned under doWrtBar && doCrdMrk (both true only when gcMode == gc_modron_wrtbar_cardmark_and_oldcheck), similarly conditioned by other callers to VMnonNullSrcWrtBarCardCheckEvaluator, and the stbx instruction generated by VMnonNullSrcWrtBarCardCheckEvaluator is conditioned under doCrdMrk = (gcMode == gc_modron_wrtbar_cardmark_and_oldcheck)

Testing: https://hyc-runtimes-jenkins.swg-devops.com/view/OpenJ9%20-%20Personal/job/Pipeline-Build-Test-Personal/29958/

[1]

 [    0x7ffee402f9f0]   15      bge     cr0, Label L4208
 [    0x7ffee402fb40]   15      lwz     gr11, [gr3, 0]
>[    0x7ffee4eb2ff0]   15      std     [gr14, 0], gr3 # #SPILL8 # SymRef  <#SPILL8_5300     0x7ffee4e542e0>[#5300  Auto] [flags 0x80000000 0x0 ]
 [    0x7ffee402fcf0]   15      ld      gr3, [gr15, 416]
 [    0x7ffee402fd90]   15      andis.  gr3, gr3, 16
 PRE: 
POST: [CCR_    0x7ffee402e640 : cr0] 
 [    0x7ffee402fec0]   15      Label L4212:    ; (Start of internal control flow)      
 [    0x7ffee402ff50]   15      beq     cr0, Label L4213
 [    0x7ffee40300b0]   15      ld      gr3, [gr15, 200]
 [    0x7ffee4030150]   15      rldicl  gr5, gr5, 0000000000000037, 007FFFFFFFFFFFFF
 [    0x7ffee4030200]   15      li      gr2, 0000000000000001
 [    0x7ffee4030350]   15      stb     [gr3, gr5], gr2
 [    0x7ffee4eb2ee0]   15      ori     gr0, gr3, 0x0
>[    0x7ffee4eb2e40]   15      ld      gr3, [gr14, 0]          ; spilled for icall # #SPILL8 # SymRef  <#SPILL8_5300     0x7ffee4e542e0>[#5300  Auto] [flags 0x80000000 0x0 ]
 [    0x7ffee40303f0]   15      Label L4213:    ; (End of internal control flow)        
 [    0x7ffee4030480]   15      subf    gr0, gr6, gr4
 [    0x7ffee4030520]   15      cmpld   cr0, gr0, gr7
 [    0x7ffee40305c0]   15      blt     cr0, Label L4208
 [    0x7ffee4030660]   15      andi.   gr0, gr11, 240
 PRE: 
POST: [CCR_    0x7ffee402e640 : cr0] 
 [    0x7ffee4030790]   15      bne     cr0, Label L4208
 [    0x7ffee4030870]   15      bl      00007FFF96D42600                ; Direct Call "jitWriteBarrierStoreGenerationalAndConcurrentMark"
 PRE: 
POST: 
 [    0x7ffee4030960]   15      Label L4209:    
 [    0x7ffee4030bc0]   15      assocreg
 PRE: 
POST: [&GPR_    0x7ffee402b060 : gr3] [GPR_    0x7ffee402aff0 : gr4] [GPR_    0x7ffee402d7f0 : gr11] [GPR_    0x7ffee4004970 : ???] [GPR_    0x7ffee402f320 : ???] [GPR_    0x7ffee402b140 : ???] [GPR_    0x7ffee402f390 : ???] [GPR_    0>
 [    0x7ffee40309f0]   15      Label L4208:    
 PRE: [&GPR_    0x7ffee402b0d0 : gr3] [&GPR_    0x7ffee402b530 : gr4] [GPR_    0x7ffee402f2b0 : gr11] [GPR_    0x7ffee402f320 : gr0] [GPR_    0x7ffee402b140 : gr29] [GPR_    0x7ffee402f390 : gr2] [GPR_    0x7ffee402f400 : gr5] [GPR_   >
POST: [&GPR_    0x7ffee402b0d0 : gr3] [&GPR_    0x7ffee402b530 : gr4] [GPR_    0x7ffee402f2b0 : gr11] [GPR_    0x7ffee402f320 : gr0] [GPR_    0x7ffee402b140 : gr29] [GPR_    0x7ffee402f390 : gr2] [GPR_    0x7ffee402f400 : gr5] [GPR_   >

Internal RTC 153007

[rmnattas]

fyi @zl-wang

[zl-wang]

Looks reasonable to me

[zl-wang]

Jenkins test sanity aix,plinux jdk8,jdk21

[zl-wang]

@AdamBrousseau looks like some infra issue for the failing build?

[zl-wang]

i would consider it as a success:

Build [Test_openjdk21_j9_sanity.openjdk_ppc64le_linux_Personal_testList_1 #196](https://openj9-jenkins.osuosl.org/job/Test_openjdk21_j9_sanity.openjdk_ppc64le_linux_Personal_testList_1/196/) completed: SUCCESS
[Pipeline] }
13:22:38  Build [Test_openjdk21_j9_sanity.openjdk_ppc64le_linux_Personal_testList_2 #196](https://openj9-jenkins.osuosl.org/job/Test_openjdk21_j9_sanity.openjdk_ppc64le_linux_Personal_testList_2/196/) completed: SUCCESS
[Pipeline] }
13:35:56  Build [Test_openjdk21_j9_sanity.openjdk_ppc64le_linux_Personal_testList_0 #196](https://openj9-jenkins.osuosl.org/job/Test_openjdk21_j9_sanity.openjdk_ppc64le_linux_Personal_testList_0/196/) completed: SUCCESS

[AdamBrousseau]

AdamBrousseau looks like some infra issue for the failing build?

Agree

[zl-wang]

all tests are successful as indicated above ... merging