|
|
@23067
|
15 years |
vboxsync |
Flush page tables that have been updated with invalid entries.
|
|
|
@23060
|
15 years |
vboxsync |
More verbose assertion
|
|
|
@22890
|
15 years |
vboxsync |
VM::cCPUs -> VM::cCpus so it matches all the other cCpus and aCpus members.
|
|
|
@22788
|
15 years |
vboxsync |
Minor pgm pool optimizations.
|
|
|
@22783
|
15 years |
vboxsync |
More stats
|
|
|
@22775
|
15 years |
vboxsync |
Statistics
|
|
|
@22774
|
15 years |
vboxsync |
Disabled experimental code that caused problems with XP
|
|
|
@22772
|
15 years |
vboxsync |
Applies to VT-x as well
|
|
|
@22771
|
15 years |
vboxsync |
Fixed raw mode regression of r51925.
|
|
|
@22767
|
15 years |
vboxsync |
Updated comment
|
|
|
@22766
|
15 years |
vboxsync |
pgmPoolFlushPageByGCPhys is currently only used in ring 3; save some …
|
|
|
@22764
|
15 years |
vboxsync |
PGMPoolFlushPage -> pgmPoolFlushPageByGCPhys
|
|
|
@22760
|
15 years |
vboxsync |
Backed out 51884 (caused gurus) and 51924 (not responsible).
Rewrote …
|
|
|
@22749
|
15 years |
vboxsync |
Check changed again
|
|
|
@22748
|
15 years |
vboxsync |
PGM pool updates
|
|
|
@22747
|
15 years |
vboxsync |
Updates
|
|
|
@22745
|
15 years |
vboxsync |
PGM pool: minor updates
|
|
|
@22742
|
15 years |
vboxsync |
paranoia
|
|
|
@22741
|
15 years |
vboxsync |
Must flush dirty pages in pgmPoolClearAll
|
|
|
@22740
|
15 years |
vboxsync |
Doesn't apply to non-paging types
|
|
|
@22738
|
15 years |
vboxsync |
Extra assertion
|
|
|
@22714
|
15 years |
vboxsync |
Removed assertion
|
|
|
@22713
|
15 years |
vboxsync |
Stats
|
|
|
@22712
|
15 years |
vboxsync |
Correction for alignment
|
|
|
@22711
|
15 years |
vboxsync |
Paranoia
|
|
|
@22710
|
15 years |
vboxsync |
Compile fixes
|
|
|
@22709
|
15 years |
vboxsync |
Also check for dirty pages
|
|
|
@22701
|
15 years |
vboxsync |
Debug checking only in VBOX_STRICT mode.
|
|
|
@22698
|
15 years |
vboxsync |
PGMPoolFlushPage must be called in an EMT.
|
|
|
@22696
|
15 years |
vboxsync |
Removed debugging code
|
|
|
@22695
|
15 years |
vboxsync |
Must flush pgm pool pages in PGMR3PhysGCPhys2CCPtrExternal to avoid …
|
|
|
@22605
|
15 years |
vboxsync |
Corrected log statement
|
|
|
@22600
|
15 years |
vboxsync |
Removed unnecessary EPT invlpg calls.
Invalidate the page who's R/W …
|
|
|
@22537
|
15 years |
vboxsync |
Another reason to flush.
|
|
|
@22533
|
15 years |
vboxsync |
Don't change the fZeroed flag.
|
|
|
@22529
|
15 years |
vboxsync |
Updated assertion
|
|
|
@22518
|
15 years |
vboxsync |
Got rid of unnecessary assertiosn
|
|
|
@22513
|
15 years |
vboxsync |
Compile fix
|
|
|
@22512
|
15 years |
vboxsync |
More verbose assertions
|
|
|
@22510
|
15 years |
vboxsync |
More statistics
|
|
|
@22507
|
15 years |
vboxsync |
Cleaned up a bit
|
|
|
@22506
|
15 years |
vboxsync |
Flush any monitored duplicates as we will disable write protection.
|
|
|
@22504
|
15 years |
vboxsync |
Updates for dirty page management (disabled code)
|
|
|
@22474
|
15 years |
vboxsync |
Compile fixes
|
|
|
@22473
|
15 years |
vboxsync |
Disabled experiment with optimized dirty PTs.
|
|
|
@22349
|
15 years |
vboxsync |
Make sure we don't kick out a page too quickly.
|
|
|
@22348
|
15 years |
vboxsync |
Tweak the defaults a bit to prevent kicking out pages too quickly
|
|
|
@22343
|
15 years |
vboxsync |
Attempt to detect full page table initialization early on.
|
|
|
@22337
|
15 years |
vboxsync |
Reset modification counter for cache pgm pool page hits.
|
|
|
@22336
|
15 years |
vboxsync |
Moved check for changed pgm pool page up.
|
|
|
@22332
|
15 years |
vboxsync |
Rex and rep prefixes set in windows 7 x64.
|
|
|
@22331
|
15 years |
vboxsync |
Deal with small memsets in 64 bits mode too (pgm pool access handler)
|
|
|
@20874
|
15 years |
vboxsync |
VMMR0CallHost -> VMMRZCallRing3[NoCpu]; VMMCALLHOST -> VMMCALLRING3.
|
|
|
@20873
|
15 years |
vboxsync |
VMMGCCallHost -> VMMRZCallRing3[NoCpu].
|
|
|
@20773
|
15 years |
vboxsync |
pgmPoolAccessHandler: Disassemble with the pgm lock taken to avoid …
|
|
|
@20764
|
15 years |
vboxsync |
More checks
|
|
|
@20763
|
15 years |
vboxsync |
Paranoia
|
|
|
@20762
|
15 years |
vboxsync |
More PGM locking
|
|
|
@20758
|
15 years |
vboxsync |
Moved assertion. Could otherwise trigger with guest SMP
|
|
|
@20748
|
15 years |
vboxsync |
PGMAllPool: assert msg.
|
|
|
@20665
|
15 years |
vboxsync |
PGMAllPool.cpp: Use VMMR3EmtRendezvous to avoid whatever issue there …
|
|
|
@20530
|
15 years |
vboxsync |
VMM: remove DISCPUSTATE from the stack.
|
|
|
@20489
|
15 years |
vboxsync |
Corrected RTLogPrintf statements and removed weird LOG_ENABLED only code
|
|
|
@20487
|
15 years |
vboxsync |
Misleading logging
|
|
|
@20374
|
16 years |
vboxsync |
*: s/RT_\(BEGIN|END\)_DECLS/RT_C_DECLS_\1/g
|
|
|
@20365
|
16 years |
vboxsync |
VMM: BEGIN_DECLS -> RT_BEGIN_DECLS; END_DECLS -> RT_END_DECLS.
|
|
|
@20151
|
16 years |
vboxsync |
Handle pgm pool flushes a bit differently.
|
|
|
@20140
|
16 years |
vboxsync |
No need to inform all VCPUs about a pending pool flush. The current …
|
|
|
@20136
|
16 years |
vboxsync |
Missing descriptions
|
|
|
@20135
|
16 years |
vboxsync |
Prevent reuse of cached larged pages with different access attributes.
|
|
|
@20058
|
16 years |
vboxsync |
Deal with pool pages being modified while we wait for the pgm lock in …
|
|
|
@19903
|
16 years |
vboxsync |
Invalidation cleanup
|
|
|
@19872
|
16 years |
vboxsync |
Cleaned up
|
|
|
@19869
|
16 years |
vboxsync |
More flushes & assertions
|
|
|
@19833
|
16 years |
vboxsync |
PGM_INVL_GUEST_TLBS -> PGM_INVL_VCPU_TLBS
|
|
|
@19808
|
16 years |
vboxsync |
Invalidate is per VCPU.
|
|
|
@19795
|
16 years |
vboxsync |
Extra assertion
|
|
|
@19794
|
16 years |
vboxsync |
Cleaned up
|
|
|
@19792
|
16 years |
vboxsync |
Atomic writes.
|
|
|
@19721
|
16 years |
vboxsync |
Locking assertions
|
|
|
@19690
|
16 years |
vboxsync |
More pgm pool locking.
|
|
|
@19680
|
16 years |
vboxsync |
More locking
|
|
|
@19675
|
16 years |
vboxsync |
Must acquire the pgm lock in pgm pool access handlers too.
|
|
|
@19674
|
16 years |
vboxsync |
Use the pgm lock during pgmPoolAlloc.
|
|
|
@19628
|
16 years |
vboxsync |
Lock counter instead of a lock flag.
|
|
|
@19627
|
16 years |
vboxsync |
PGM pool: allow multiple users of top-level paging structures …
|
|
|
@19572
|
16 years |
vboxsync |
Started with ring 0 stack trace dump support.
|
|
|
@19516
|
16 years |
vboxsync |
Always call pgmPoolSyncCR3 from PGMSyncCR3 (regardless of guest paging …
|
|
|
@19503
|
16 years |
vboxsync |
Fixed r47022 regression
|
|
|
@19488
|
16 years |
vboxsync |
Made pgmPoolSyncCR3 a bit more SMP safe (pgmPoolClearAll still to be done).
|
|
|
@19330
|
16 years |
vboxsync |
Cleaned up a bit
|
|
|
@19329
|
16 years |
vboxsync |
Removed pgmPoolGetPageByHCPhys.
|
|
|
@19319
|
16 years |
vboxsync |
More verbose assertion
|
|
|
@19141
|
16 years |
vboxsync |
Action flags breakup.
Fixed PGM saved state loading of 2.2.2 images. …
|
|
|
@19024
|
16 years |
vboxsync |
PGM: Moved shadow mode exit and re-enter from PGMAllPool.cpp to …
|
|
|
@18992
|
16 years |
vboxsync |
More PGM api changes
|
|
|
@18988
|
16 years |
vboxsync |
PGM api changes
|
|
|
@18986
|
16 years |
vboxsync |
Fixed virtual handler update regression.
|
|
|
@18984
|
16 years |
vboxsync |
Autoset does belong in PGMCPU
|
|
|
@18982
|
16 years |
vboxsync |
Compile fix
|
|
|