46ae097c942b4a2d5038d9593e312856238da75f 46420 |
|
06-Jun-2013 |
vboxsync |
VMM, recompiler: Purge deprecated macros. |
a3cf6fbb13cc7700ad8b080f16eb3e11efcd8239 45836 |
|
30-Apr-2013 |
vboxsync |
Spelling. |
28253aaeecd3b941f68e0253a93062bd19523912 45826 |
|
30-Apr-2013 |
vboxsync |
PGM: Extended the WP0+RO+US hack to include big pages and adjusted some assertions to handle the hack. |
bd960df9d97611663b270c6a7b91df40d64509b4 45798 |
|
29-Apr-2013 |
vboxsync |
Fixed up and enabled Netware WP0+RO+US hack. |
e88cd082acdb680fee99ecf34d45c4fa261bcfaa 45797 |
|
29-Apr-2013 |
vboxsync |
PGMAllBth.h: Duh, not >, but <, darn it. |
904e375501473bc9ac7acd1531421b41c1a8cb09 45792 |
|
28-Apr-2013 |
vboxsync |
CR0.WP applies to CPL > 3, not just CPL == 0. Sketched out a Netware hack. |
c3e58ec2ac4809e342d11ea7dd7d873b7b97ed1f 45752 |
|
26-Apr-2013 |
vboxsync |
Don't call MMHyperIsInsideArea if we're using HM to execute code, it will return bogus results! |
4f3c9f17b22e19115e16fd1a9ee2eb62d5ef4e20 45735 |
|
25-Apr-2013 |
vboxsync |
VMM: Consistently flush TLB entries for non-present pages. |
ffd8f69bceafdc5da6cd24070a30ed8eb1df8e2a 45619 |
|
18-Apr-2013 |
vboxsync |
PGMAllBth.h: Skip a path in Trap0eHandler that hooks into CSAM if CSAM is disabled, avoid some work (like disassembling the instruction). |
30d260820bc5dfc0d147b843740c252fde2e699d 45407 |
|
08-Apr-2013 |
vboxsync |
VMM: Invalidate page even when it's not present now. |
7ce6e7e8fb0eddb176361a49f53fa1bd15eaab4e 45276 |
|
02-Apr-2013 |
vboxsync |
Ring-1 compression patches, courtesy of trivirt AG:
- main: diff to remove the hwvirt requirement for QNX
- rem: diff for dealing with raw ring 0/1 selectors and general changes to allowed guest execution states
- vmm: changes for using the guest's TSS selector index as our hypervisor TSS selector (makes str safe) (VBOX_WITH_SAFE_STR )
- vmm: changes for dealing with guest ring 1 code (VBOX_WITH_RAW_RING1)
- vmm: change to emulate smsw in RC/R0 (QNX uses this old style instruction a lot so going to qemu for emulation is very expensive)
- vmm: change (hack) to kick out patm virtual handlers in case they conflict with guest GDT/TSS write monitors; we should allow multiple handlers per page, but that change would be rather invasive |
40642fb7f08e1b6c55284d571b47c06d103514b2 45103 |
|
20-Mar-2013 |
vboxsync |
PGMPool: Eliminated fictive page (except NIL). #6367 |
c58f1213e628a545081c70e26c6b67a841cff880 44528 |
|
04-Feb-2013 |
vboxsync |
header (C) fixes |
98427c0ab08697e468c26dc33ee9571308577867 44399 |
|
27-Jan-2013 |
vboxsync |
DBGF,DBGC,++: PVM -> PUVM. Some refactoring and cleanup as well. |
f091ce66ee934d599f16056078a9a76d7286b959 43387 |
|
21-Sep-2012 |
vboxsync |
VMM: HM cleanup. |
201339bee6d3364b123b1e73dd9869f415f8ae12 42188 |
|
17-Jul-2012 |
vboxsync |
VMM: Changed a few ifndef IN_RING0 to ifndef VBOX_WITH_RAW_MODE_NOT_R0. |
681380880d4131019871e8f22cb3349b757168ca 42186 |
|
17-Jul-2012 |
vboxsync |
SELM,DIS,CPUM,EM: Hidden selector register cleanups. |
7a896688c49bde3fa1490e7ebb321ac51b6ad29d 41939 |
|
28-Jun-2012 |
vboxsync |
CPUMGetGuestCPL: Drop the context core pointer and use the Guest state in CPUMCPU via pVCpu. |
37fb67be7d1d328213aeda3f56ab5aacd37416d1 41906 |
|
24-Jun-2012 |
vboxsync |
CPUM: Combined the visible and hidden selector register data into one structure. Preparing for lazily resolving+caching of hidden registers in raw-mode. |
67b4b089c50d0ab8ad847dddd8c0e0812fdadc9e 41836 |
|
19-Jun-2012 |
vboxsync |
Doxygen. |
df45e3751b773c12f2c0838f63500a50cf535e86 41823 |
|
19-Jun-2012 |
vboxsync |
Avoid using SELMValidateAndConvertCSAddr+EMInterpretDisasOneEx when possible and call EMInterpretDisasOne instead. Changed EMInterpretDisasOne to use SELMValidateAndConvertCSAddr instead of SELMToFlatEx. |
df25990f935e7fd32acd9be9a156aff8d10facf2 41802 |
|
17-Jun-2012 |
vboxsync |
Doxygen. |
4bfa7b58e362a1bca0628643c352c137900bf01a 41783 |
|
16-Jun-2012 |
vboxsync |
Doxygen, comment typos. |
5981e6935987b08737b730b63a41acc1dd696377 41774 |
|
16-Jun-2012 |
vboxsync |
bugref.. |
98502a585f6eda30527b54f7df18dd3de3d3d7c6 41675 |
|
12-Jun-2012 |
vboxsync |
DIS: The start of the scoping exercise... Fixed PREFIX_ADDRSIZE that I just busted in the previous commit. |
0579aa7698cf542ea9efd5f515e822d553a7d648 41458 |
|
28-May-2012 |
vboxsync |
PGMPool: Adding A20 state to the page pool (partly disabled). |
541071eef3db0e3e0e5497bb4b692efee42b1ad0 41456 |
|
28-May-2012 |
vboxsync |
PGMPool: PGMPOOLPAGE adjustments and cleanups, no real code changes intended. |
cbf70cda8a0d6e49304ef3c9ff6fa77f71606fc6 41420 |
|
23-May-2012 |
vboxsync |
PGM: More A20 fixes. |
e7287801140c21acfa5ddd76668921e7dfed49eb 41391 |
|
22-May-2012 |
vboxsync |
PGM: A quick stab at correct A20 gate masking (new code is disabled). |
d521d3c5e4fa92bb76452c04cad2d1e5bf0d50ed 40449 |
|
13-Mar-2012 |
vboxsync |
SELM: Refactoring (PVM -> PVMCPU). |
ceec4276710e550e3d28840e12501f4da1e8ad7b 40274 |
|
28-Feb-2012 |
vboxsync |
Introduced VBOX_WITH_REM in Config.kmk and the VMM. |
1cee4e2bc3fb261ec5e68aaae1377cdb8fe82849 39908 |
|
30-Jan-2012 |
vboxsync |
Disabled seemingly harmless assertion. |
683371bbf37760161d1b8454ce978acf89bbb04f 39402 |
|
23-Nov-2011 |
vboxsync |
VMM: don't use generic IPE status codes, use specific ones. Part 1. |
fa5a0e7d83e683124691e0b3fa14e40d4ab6ea8b 39301 |
|
15-Nov-2011 |
vboxsync |
PGMAllBth.h: Disabled annoying SMP assertions. |
215ca497e5248c6d92dcf11ba4ef9d000ab471e2 39284 |
|
11-Nov-2011 |
vboxsync |
PGMAllBth.h: disabled annoying assertions. |
2afbe132eb7931e0125141eabe3a48e08f1ffab5 39078 |
|
21-Oct-2011 |
vboxsync |
VMM: -Wunused-parameter |
d03c6bcd3d78cef8ac4d76186e7c0e72f63ce80f 39038 |
|
19-Oct-2011 |
vboxsync |
VMM: -W4 warnings (MSC). |
d2c6b2e8826a5ef34170fef0c72c3fc7c5c1b46a 39034 |
|
19-Oct-2011 |
vboxsync |
VMM,INTNET: Addressing unused variable warnings. |
c7a00ac75c7941df2afb62e6fd7ffdf1795e6c76 38953 |
|
06-Oct-2011 |
vboxsync |
PGM: Attempt at fixing the VERR_MAP_FAILED during state save problem on 32-bit hosts when assigning lots of memory to the guest. PGM should lock down guest RAM pages before use and release them afterwards like everyone else. Still quite some stuff left to do there, so I've deviced a little hack for tracking unlocked mappings and using this as input when deciding to do async or sync chunk unmapping at save/load time. See xtracker #5912 and public ticket 7929. |
fa79e346ae60eee8b82eb6c5da64ab2c3a76b5f7 38707 |
|
09-Sep-2011 |
vboxsync |
VMM/VT-x: Fix for PAE guests running on 32-bit hosts or 64-bit hosts where VBoxInternal/PGM/MaxRing3Chunks is used. |
de6c675bba7ec5e13f9f80900296530482fde8ba 38086 |
|
21-Jul-2011 |
vboxsync |
oooops. |
3b3bc8a9383a065307e540b83fc3a3d6c548a082 38082 |
|
20-Jul-2011 |
vboxsync |
*.kmk: Installation/staging adjustments (kBuild 0.2.0 preps). |
addc480d0d7650db6323467bbdab6c21836a2928 37354 |
|
07-Jun-2011 |
vboxsync |
PGM: Fixed locking issues in PGMR3PhysMMIORegister and PGMR3PhysMMIODeregister. Also addressed a harmless on in PGMR3PhysRomRegister (only used at init time, so no races). Fortified the code with assertions more lock assertion, replacing the incorrect PGMIsLocked() checks (we only care if the current thread is the lock owner). Cleaned up some ReturnStmt macros and adding more of them. |
1999ae03c34840fa4d712fd2e020120b2cb7182a 36891 |
|
29-Apr-2011 |
vboxsync |
PGM: Put a TLB in front of the RAM ranges to speed up lookup (disabled). |
7b5d0a97e9f87008f7b31583653d93dccaffe5ce 36009 |
|
17-Feb-2011 |
vboxsync |
PGM: Fixed large pages and write monitoring (live snapshot). Added checks for PGM_PAGE_PDE_TYPE_PDE_DISABLED in a few places where only PGM_PAGE_PDE_TYPE_PDE was checked for (might have missed some). |
4b2dbdac954a68bf8bd20dc33ba98774e3fa013c 35754 |
|
28-Jan-2011 |
vboxsync |
Fix running 64bit guests on 32bit hosts with nested paging |
1f2c889c173f5253ef9655d6cd1017f3868b1f9b 35537 |
|
13-Jan-2011 |
vboxsync |
mark the source of the bloat. |
b791c122cd86e4f2fcd2cb13aa7e48b8421b1876 35529 |
|
13-Jan-2011 |
vboxsync |
warnings (gcc-4.5). |
6b1b86b85d9a526e354a02460dcf21113c943d5c 33795 |
|
05-Nov-2010 |
vboxsync |
Extra comment |
bc1484a141a5638d1c26739e77e8a47c77dc2da3 33676 |
|
02-Nov-2010 |
vboxsync |
scm cleanup run. |
ad27e1d5e48ca41245120c331cc88b50464813ce 33540 |
|
28-Oct-2010 |
vboxsync |
*: spelling fixes, thanks Timeless! |
ac68e451413f09dada5815439a3b212331bda3d3 32788 |
|
28-Sep-2010 |
vboxsync |
Moved stats around |
8b9a1eee7e54b84aa9e2fa1a43831e7a0c96782e 32785 |
|
28-Sep-2010 |
vboxsync |
Cleaned up SyncPageWorker for non-paging modes. |
53a589f6732d9b92d7409c9cb0fc4cee353a960f 32778 |
|
27-Sep-2010 |
vboxsync |
Sync the entire page table for new nested page tables. (partially syncing them doesn't really make sense) |
dc280865bd72003e72d83a80fa88275dbb27585f 32564 |
|
16-Sep-2010 |
vboxsync |
Compile error |
4d7c8080ba94df6e023b1fb9c1c89df0964823c2 32563 |
|
16-Sep-2010 |
vboxsync |
Take a shortcut |
a6e74ed4a42ad70663550b898b45049d4abb0328 32561 |
|
16-Sep-2010 |
vboxsync |
Oops |
d3a983be073f0d3ae5b64839c10be142bc0b8970 32560 |
|
16-Sep-2010 |
vboxsync |
Cleaned up a bit |
1ca3e31477ecb518c4866dff5a3cae0d043467a5 32541 |
|
16-Sep-2010 |
vboxsync |
Don't bother checking for virtual handlers in ring-0. |
1a2cdb8d0dffa9bf649e3610d7d39c01af8725c1 32539 |
|
16-Sep-2010 |
vboxsync |
Removed wrong assertion |
d89dccbab8cc69c1bde9cbc748c52457ec221783 32521 |
|
15-Sep-2010 |
vboxsync |
Optimization |
542f1852882c92fc5a13ec5819b68cbac6c9d1b5 32486 |
|
14-Sep-2010 |
vboxsync |
Moved logging and stat |
6f6d7a02a4e211aec2a8180c91582d2d84756894 32480 |
|
14-Sep-2010 |
vboxsync |
SyncPageWorkerTrackDeref: Use the hint we retrieved from the cached dirty guest PT.
Do not reset the modifications counter in InvalidatePage. |
b35e3948f1287430503b6b432945b8cf4bfd3a23 32431 |
|
11-Sep-2010 |
vboxsync |
scm cleanup |
20d378173c586e3def8fe1aadaa4c22b956c84e1 32410 |
|
10-Sep-2010 |
vboxsync |
Got rid of unnecessary dirty page flush in InvalidatePage. (NOTE: high risk change) |
086e613371e0f6d1efc38fbc199fa20291afcd5e 32401 |
|
10-Sep-2010 |
vboxsync |
Updates |
6c7440ab575cca1aa4d97d96b2d898a0d20c8efc 32384 |
|
10-Sep-2010 |
vboxsync |
Extended dirty page optimization for the pae/32-bit shw/gst combination (disabled). |
2ac7a0ca33ebd20f2ee34cbf7246cd49ee6a5dd4 32366 |
|
09-Sep-2010 |
vboxsync |
Typo |
ef481e64bafc49f1232767ce12c3fa69a5d5fe50 32362 |
|
09-Sep-2010 |
vboxsync |
Some cleanup |
4a8c9f8323eb7c9760091ee0e95cfc332614ad08 32042 |
|
27-Aug-2010 |
vboxsync |
PGMAllBth.h: MMIO opt fix. |
b40179b44fea65b72b2f226f62af1ed7bd3c48fc 32036 |
|
27-Aug-2010 |
vboxsync |
Removed X86_PTE_PAE_PG_MASK, renamed X86_PTE_PAE_PG_MASK_FULL to X86_PTE_PAE_PG_MASK. |
4a33af43d9d1a46f9c239af3fc8ce6d72f289538 32035 |
|
27-Aug-2010 |
vboxsync |
Changing the remaining X86_PTE_PAE_PG_MASK uses to X86_PTE_PAE_PG_MASK_FULL. |
a7b04151fe55788dd0da05e421f35d82b8381bf5 32028 |
|
27-Aug-2010 |
vboxsync |
PGMAllBth/SyncPage: Don't assert on stale TLB after W7-64 changed some A bits. Just invalidate the entry and try again. |
c5c8d1e0169f6f7e4439e7325a0c61ce3ecab3bc 32027 |
|
27-Aug-2010 |
vboxsync |
PGM/SyncPage: To be on the safe side, always sync the target page when doing multiple pages. |
9f9a20823b87e89c1b5cb45eb9b5699b29bfefeb 32009 |
|
26-Aug-2010 |
vboxsync |
Removed GST_PDPE_PG_MASK and renamed GST_PDPE_PG_MASK_FULL to GST_PDPE_PG_MASK. |
9ccc5e33c3f10daa33f4da258e0f2d555b7dfe0f 32005 |
|
26-Aug-2010 |
vboxsync |
Changed all X86_PDPE_PG_MASK references to X86_PDPE_PG_MASK_FULL. |
9673b0b804d0b0f801f377c6371600e92071241f 32000 |
|
26-Aug-2010 |
vboxsync |
x86.h,*: Eliminated X86_PML4E_PG_MASK and renamed X86_PML4E_PG_MASK_FULL to X86_PML4E_PG_MASK. |
5995ea736700c896d73e686a80d87f0ad3dcc08d 31998 |
|
26-Aug-2010 |
vboxsync |
Changed the remaining X86_PML4E_PG_MASK uses to X86_PML4E_PG_MASK_FULL. |
e4a93d2f3ca0b165b618c27d37d984b05d24bed8 31996 |
|
26-Aug-2010 |
vboxsync |
Mostly safe X86_PTE_PAE_PG_MASK -> X86_PTE_PAE_PG_MASK_FULL conversions. |
7b0d0df4cf38fe9704783dc7cc82aeec79ca4df5 31994 |
|
26-Aug-2010 |
vboxsync |
Safe X86_PDPE_PG_MASK -> X86_PDPE_PG_MASK_FULL conversions. |
ff2ff83bf0faae1171ab33f9a57d785b44618f11 31989 |
|
26-Aug-2010 |
vboxsync |
VMM: Removed the PGMR3DumpHierarchyGC/HC functions. |
f7ab406533db8025dde58fffbe4669e7d9cb381d 31947 |
|
25-Aug-2010 |
vboxsync |
PGM: paging hierarchy dumpers refactoring in progress. |
1e3cd9c9c95ff352da1331a1789b05fd012c880d 31870 |
|
23-Aug-2010 |
vboxsync |
PGM: Don't sync unnecessary PTE bits. |
a5c9bd1199e10d51c8f5fef6293b765845ada754 31851 |
|
23-Aug-2010 |
vboxsync |
pgmPoolAllocEx/pgmPoolAlloc: Just use pgmPoolAllocEx when you want locking. |
a5c11842b92303b7b5d0086ee5011a969e321b50 31849 |
|
22-Aug-2010 |
vboxsync |
PGM: Don't shadow PDE bits 62/3:52 (NXE=1/0), only shadow the bits we need to (exception G, A and D). |
de6e197433bb41d766a6bf3febd32f5234f74268 31835 |
|
21-Aug-2010 |
vboxsync |
PGMAllBth.h: fixed assertion. |
2998d42c13fb82e9e2a1268f3c7b1047d5a6d7d6 31832 |
|
21-Aug-2010 |
vboxsync |
PGM: The other w7-64 regression. Trouble is that they're using PTE bits 52:62 for something. |
1fadd17f5f9d7bfcd5770da325b481c7334929ac 31823 |
|
20-Aug-2010 |
vboxsync |
nc |
375cb0b594a9101c1e9460219836931aa1a854a9 31807 |
|
20-Aug-2010 |
vboxsync |
FT: disable large page usage
Removed debug code |
7606410fd1d01ccde408fce119caffc40ad6dfe8 31784 |
|
19-Aug-2010 |
vboxsync |
PGM: Make sure we deploy the MMIO hack on all possible pages. Physical handler stats. |
bdcc3bc6e65e18faed46372067205be03489088c 31780 |
|
19-Aug-2010 |
vboxsync |
PGM: Addressed a fixme. logging and minor cleanup. |
2634ec5cbf8f1fa0a968cd4664ead6df1fed730d 31775 |
|
19-Aug-2010 |
vboxsync |
PGM: Wrap up all access to PAE/LM PTEs so that we can treat the invalid entries used by PGM_WITH_MMIO_OPTIMIZATIONS as not-present. |
020916e230faa4a37bace3a8480f54fac94ea173 31657 |
|
13-Aug-2010 |
vboxsync |
PGMPool: Drop cRefs since it's always 1. |
f351fab06d1e1507b72f5c574df71193061fd393 31636 |
|
13-Aug-2010 |
vboxsync |
EM: Addressed VBOXSTRICTRC todo from r64673. Decided to only use VBOXSTRICTRC on the methods that would actually return VINF_EM_* stuff. |
f7c77958d7da7c76cb2ff6e07eae79399d305408 31612 |
|
12-Aug-2010 |
vboxsync |
Compile fix |
eadf3092d11e2c9acd44daf18e361d18a1f8416b 31610 |
|
12-Aug-2010 |
vboxsync |
Wrong check |
8746d1785c98864e4db846647c03f73e48107035 31600 |
|
12-Aug-2010 |
vboxsync |
Extra debug command |
3f1e0eea71cabeb90529e546f16eb7aee513fde9 31593 |
|
12-Aug-2010 |
vboxsync |
PGM,IOM: MMIO optimization - hacking in progress. (still disabled) |
cc723cf07e365cd40b517b9c5da4f113e9469745 31565 |
|
11-Aug-2010 |
vboxsync |
PGM,HWACCM: MMIO optimizations - coding in progress. |
c2fecdb51a36bf219484cd0068294d49060d78d0 31447 |
|
07-Aug-2010 |
vboxsync |
InvalidatePage: Drop the unused hint here as we'll be releaseing the AutoSet shortly after returning anyway. |
821c004ccf9508ae9c4d51a43e90fa0f77ddf76c 31443 |
|
06-Aug-2010 |
vboxsync |
PGMAllBth.h: Must pass the RW flag to SyncHandlerPte. Some more logging. |
9984bae4ef25be8afd95d184308adcfef6e2ec0d 31402 |
|
05-Aug-2010 |
vboxsync |
PGM: Replaced the hazzardous raw-mode context dynamic mapping code with the PGMR0DynMap code used by darwin/x86. This is a risky change but it should pay off once stable by providing 100% certainty that dynamically mapped pages aren't resued behind our back (this has been observed in seemingly benign code paths recently). |
7727a3247478ea62d7070c51365c9f0d17224fd2 31207 |
|
29-Jul-2010 |
vboxsync |
PGMAllBth.h: Temporarily shut up ugly raw-mode context assertion. |
c5a3a5673282237d1b8bfc6b22659d42378ce283 31178 |
|
28-Jul-2010 |
vboxsync |
PGM: Some more preparatory cleanup. |
25d9ca299a26aa8a5f6ee6e7c60c66d189a9b6f4 31170 |
|
28-Jul-2010 |
vboxsync |
PGM: More inline cleanup (prep for tracking dynamic mappings in raw-mode). |
4f2ee5b4e2c440c55272059f1f8686d8d29c4244 31167 |
|
28-Jul-2010 |
vboxsync |
PGM: Pass pVCpu instead of pPGM (PPGMCPU) and eliminate PGMPOOL_PAGE_2_PTR_BY_PGM and PGMPOOL_PAGE_2_PTR_BY_PGMCPU. |
67a1f4df3696efa09203887f06e3d6602d544daf 31141 |
|
27-Jul-2010 |
vboxsync |
PGM: If we get to the end of Trap0eHandler it doesn't mean it is a guest fault, but that RT_FAILURE(rc). ASSUMING this always means that it is time for a guru meditation. |
1c0fc07b1899cffc3cbb168020f4153533141865 31140 |
|
27-Jul-2010 |
vboxsync |
PGM: Removed CheckPageFault as it is no longer used. Removed pfn*SyncPage since it is always used directly. Made a few more internal functions static to encourage compiler optimizations. |
def71a910acf99c885592fd35b2892d2d20b2320 31136 |
|
27-Jul-2010 |
vboxsync |
PGM: cache the last physical handler lookup result in each ring. |
1986f56777969a25707ab214f8dd070804be666c 31123 |
|
26-Jul-2010 |
vboxsync |
PGM: Moved the PGM stats out of the VM structure. Reduces the size of the PGMCPU structure quite a bit. |
a3019da27e24031fff359d48ddd5aa598ce9480e 31119 |
|
26-Jul-2010 |
vboxsync |
docs |
93cd29d200c115b23fb13ab3b2e57c92aa628a80 31102 |
|
26-Jul-2010 |
vboxsync |
PGM: #PF stats. Drop the old ones in favor of StatRZTrap0eTime2*. Added a few new stats to that group. |
acdffbabea85512922d7130ec8ff671772f926cc 31101 |
|
26-Jul-2010 |
vboxsync |
PGM: Moved the code dealing with access handlers out of the #PF function. |
bf2c812cbf1a839868d71cb9619694b8aab2db0b 31093 |
|
26-Jul-2010 |
vboxsync |
PGM: More #PF cleanup, eliminating a few PGMGstGetPage() calls that are unnecessary now. |
a331c988bd55b3fb0b20b204f2bdc24a7cf16354 31092 |
|
26-Jul-2010 |
vboxsync |
PGM: Some #PF cleanup, no code change just rearranging and unindenting it a bit. |
93b426c62359d7b684b743e9ae959f137ae1b79c 31090 |
|
26-Jul-2010 |
vboxsync |
PGMAllBth.h: Cleaned out the old code. |
8d13f414cc4b457062c0b7b68e4ac7363684b96a 31089 |
|
26-Jul-2010 |
vboxsync |
PGMAllBth.h: Enabled the GstWalk code again as it doesn't seem to make any difference on the tests after all. |
13ac1dfa191ffa78eecf75d963827ff44025b78b 31086 |
|
25-Jul-2010 |
vboxsync |
PGM: Disabled the GstWalk #PF code. |
5fe5272ab0baa8499b05e0534c37f85a80d8cd99 31082 |
|
24-Jul-2010 |
vboxsync |
build fix? |
a37a9e3ee7a31e899256cb5d94c67159fab22cdc 31081 |
|
24-Jul-2010 |
vboxsync |
PGM: Use GstWalk in the #PF, part 1. |
f163417381bbf0521a5cd6ab6f25a4417b5ecfcd 31069 |
|
23-Jul-2010 |
vboxsync |
PGM: Prep for shadowing EFER.NXE and CR4.PSE to avoid function calls. |
05368cb8dfc9dd3170546ffda37f0072fc31fb38 31067 |
|
23-Jul-2010 |
vboxsync |
PGM: A couple of simplifications and optimizations. -assertions |
1718098fb08fdb7dafd85de1aaf51debc9e72828 31066 |
|
23-Jul-2010 |
vboxsync |
PGM: A couple of simplifications and optimizations. |
d0052e6f3561b4204cf99b2fdb2df8a1c06ae5f7 30895 |
|
17-Jul-2010 |
vboxsync |
build fixes. |
2e0f39c144cbad4d4e8d18695334a29653240e26 30891 |
|
17-Jul-2010 |
vboxsync |
PGM: Cleanups related to pending MMIO/#PF optimizations. Risky. |
7c9a5eca233baf6ede345ace077a00bd0b7af1ef 30889 |
|
17-Jul-2010 |
vboxsync |
PGM: Cleanups related to pending MMIO/#PF optimizations. Risky. |
34f9eef373e9fe42afc6ed33b484afbaa4fd7b73 30325 |
|
21-Jun-2010 |
vboxsync |
Obsolete pgmLock/Unlock |
d3adf5695782fe54c34d3436cb453a5620ee86fc 30263 |
|
16-Jun-2010 |
vboxsync |
VMM,REM: Only invalidate hidden registers when using raw-mode. Fixes save restore during mode switching code like the windows boot menu. (#5057) |
2077277ac48e82c339aba092cdb89f1f2010afbf 29676 |
|
20-May-2010 |
vboxsync |
Too noisy |
bc0229fce4e9f3a2938f0e8c680c3d92f78351dc 29217 |
|
07-May-2010 |
vboxsync |
Shared paging updates |
e64031e20c39650a7bc902a3e1aba613b9415dee 28800 |
|
27-Apr-2010 |
vboxsync |
Automated rebranding to Oracle copyright/license strings via filemuncher |
d8aefbf86d1d2db659ee87d2d2d58f3a1b14e1e9 28753 |
|
26-Apr-2010 |
vboxsync |
Don't set X86_PDPE_NX in the fake PML4E or PDPTE. |
c2923d82c1b72d03e2a556a50b9708302d1e66ad 28750 |
|
26-Apr-2010 |
vboxsync |
Mask away the page offset |
82e90599291da476b2de7c8db33cfb0f2cbac774 28656 |
|
23-Apr-2010 |
vboxsync |
pgmPoolTrackPhysExtDerefGCPhys needs to check the PTE index as well in order not to kick out the wrong slot |
5007ad5d2bce84d88ee339a94e0aa7b91c699194 28639 |
|
23-Apr-2010 |
vboxsync |
Keep track of page table entry indices and deal with multiple physical page references in one page table. |
ba6e3362790a1ec580c4ffa0734af2fec0c0f4c6 28458 |
|
19-Apr-2010 |
vboxsync |
Cleared entire page table entry |
e0421ee4efb1bbf81bc4740b530fc245aed8152f 28123 |
|
09-Apr-2010 |
vboxsync |
Simplified hypervisor conflict check |
89f2c85d45275b63e62ae8c95421e3668cfba7c2 28094 |
|
08-Apr-2010 |
vboxsync |
Compile fix |
1641591fedc000a32c927fcdcf0b21606d36af0b 28090 |
|
08-Apr-2010 |
vboxsync |
Check for write conflicts with our hypervisor mapping early on. If the guest happens to access a non-present page,
where our hypervisor is currently mapped, then we'll create a #PF storm in the guest. |
468c2bcb36eb9a032f5dd0fcb34db10bd58e9996 27976 |
|
04-Apr-2010 |
vboxsync |
*: scm cleans up whitespace and adds a new line at the end of ApplianceimplPrivate.h. |
5871b462154c3deb60db077635ca01db7fd94175 27814 |
|
30-Mar-2010 |
vboxsync |
Backed out part of r59436 |
cf262e3e40021064eeeb7a189bae5cbac5265f14 27790 |
|
29-Mar-2010 |
vboxsync |
Cleaned up |
ba6f599ede8e91d5ea1fd01e6b98d5b055501e94 27783 |
|
29-Mar-2010 |
vboxsync |
PGMSyncCR3: pgmPoolResetDirtyPages must be called before pgmPoolSyncCR3 |
6d3f37f74ec27da375561640b983e165a55bb9b3 27584 |
|
22-Mar-2010 |
vboxsync |
Statistic counter for ballooned page accesses |
6e932c9bc474367bcdb27fa9235f72e46711f091 27583 |
|
22-Mar-2010 |
vboxsync |
Emulate reads from ballooned pages as they are not present in our shadow page tables. (required for e.g. Solaris guests; soft ecc, random nr generator) |
549a53f649248476f7c4abfd463a82e958abc8ec 27540 |
|
19-Mar-2010 |
vboxsync |
Step one in dealing with ballooned pages in shadow paging |
c1f37b8a9684ba71203de5778d9caf747c3863b2 27490 |
|
18-Mar-2010 |
vboxsync |
More comment + paranoid assertions |
3568e65c6e903af44a117fd330d14c23440b7e40 27489 |
|
18-Mar-2010 |
vboxsync |
Disabled the lazy syncing of zero pages based on the u1Write setting. Currently causes inconsistencies with PAE dirty pages
and looks quite inefficient (flushing if multiple references). |
bd896de5b4d566c513888905a1daeea3ba74290a 27488 |
|
18-Mar-2010 |
vboxsync |
No need for pgmPhysPageMakeWritableUnlocked |
ff295e43b74d8278aea6ca1b50948cff8cee1f0a 27486 |
|
18-Mar-2010 |
vboxsync |
Style |
0e363b45a38c651285894793becb5903bf182237 27210 |
|
09-Mar-2010 |
vboxsync |
Note added |
be6c654770a02f5eaf8d4cd59d766034f6bb59f4 27209 |
|
09-Mar-2010 |
vboxsync |
Backed out r58570; must decrease reference here |
73309e274785dc8266f02e373167bb25cc4ae1d8 27203 |
|
09-Mar-2010 |
vboxsync |
Double derefence removed |
6d98aee0da4bf4d82ea936a7aea7f2e3bc9ecbd3 27113 |
|
05-Mar-2010 |
vboxsync |
Logging |
9c9a35cf750af1c7ea049e1a8040bf8faeecad7d 27112 |
|
05-Mar-2010 |
vboxsync |
Missing u1User bit for AMD-V large pages |
e68478f375e2facbf5601ebd59d060aace950084 27110 |
|
05-Mar-2010 |
vboxsync |
Logging update |
059bfdd0f02b6881186c008ed1e1ad1f7c39d484 27108 |
|
05-Mar-2010 |
vboxsync |
Logging updates |
b301c2ed80b2ae5201801c30b37618a158fcc14e 27038 |
|
04-Mar-2010 |
vboxsync |
Try to reuse a 2mb large page if it was previously disabled for monitoring purposes. |
fc4516250f652da12d477446bc5351bcf6c5aaeb 27026 |
|
04-Mar-2010 |
vboxsync |
More large page work |
05d752e82261cfe829ddbc31aa3c7bc681a8a5b7 26849 |
|
26-Feb-2010 |
vboxsync |
-Wshadow burns |
941296ca8d34c7656018d52ddf11d0105302b398 26841 |
|
26-Feb-2010 |
vboxsync |
AMD-V support for large pages (64 bits host, nested paging) |
78c4c0eefd6d0871a97f14ff7a0b9d0e5a840593 26727 |
|
24-Feb-2010 |
vboxsync |
Stats |
0cf941aff08f08fba814e9ec6b95c99c8a5166b2 26717 |
|
23-Feb-2010 |
vboxsync |
Large page cleanup |
d995b8126ec969ca3ffcaf0e2393a3f33f64e5c9 26711 |
|
23-Feb-2010 |
vboxsync |
Moved check up |
4b368ef9be46cbf2b0f3ae67243503675e2249a7 26701 |
|
23-Feb-2010 |
vboxsync |
Paranoia & logging |
5eff5730c29c081e04cfd1e0468abbf9fde98851 26700 |
|
23-Feb-2010 |
vboxsync |
Deal with missing PDEs in SyncPage for SMP guests.
Deal with big PDEs in SyncPage for SMP guests (nested paging only). |
a2f3168423b7487970bdd584c86a7a911c5ff4d6 26685 |
|
22-Feb-2010 |
vboxsync |
Large page changes |
740c481388d9b3c9ced757bedba9786acd8c24d6 26635 |
|
18-Feb-2010 |
vboxsync |
PAE and AMD64 paging modes support large pages regardless of CR4.PSE. |
89e315209cb9fb92729218ba5c0a5bc3d83e9840 26625 |
|
18-Feb-2010 |
vboxsync |
Large page code cleanup |
d8d39167a811c992da9398837b002676682783b6 26621 |
|
17-Feb-2010 |
vboxsync |
Some counters |
aab737754188eb6b8cef9b8553c147e0642d49a1 26620 |
|
17-Feb-2010 |
vboxsync |
Wrong mask |
a2109ad90c6803bd75327c1e08bff506fadc57d0 26619 |
|
17-Feb-2010 |
vboxsync |
Logging |
23317edd1954beaac3087584b2bc09fac3e26822 26616 |
|
17-Feb-2010 |
vboxsync |
More large page work (disabled) |
5acd506d7f80295c8328715f744117fddb0b3444 26577 |
|
16-Feb-2010 |
vboxsync |
More large page work |
45787b7d73c05dc8e5f4f91239ee2bbac3cb169e 26570 |
|
16-Feb-2010 |
vboxsync |
PGM lock has already been taken by caller |
2b413546a8817745056f51692807076ba1d0eb5b 26303 |
|
05-Feb-2010 |
vboxsync |
PGMAllBth.h: warnings |
a46e9b10272ca13974341175f4f850a51f8887fb 26278 |
|
05-Feb-2010 |
vboxsync |
VMM: More warnings. |
5fb478ec53221ea324bb866728dabecb8ffa24b1 26277 |
|
05-Feb-2010 |
vboxsync |
VMM: more RC/GC warnings - CSAMDoesPageNeedScanning and CSAMMarkPage now takes RTRCUINTPTR instead of RTRCPTR to ease the pain using them in RC. |
1f19ba1642702523c0d2684aa2542d051f15275b 26271 |
|
05-Feb-2010 |
vboxsync |
VMM: warnings. Changed PATMIsPatchGCAddr and CSAMIsKnownDangerousInstr to take RTRCUINTPTR instead of RTRCPTR so we can mostly avoid having to cast the parameter. |
5463c845e3d7ce94adc0474c5d8de30dbb931bad 26218 |
|
03-Feb-2010 |
vboxsync |
*.kmk: Started renaming templated from all caps to CamelCase. At the same time preparing to make warnings errors (VBOX_WITH_WARNINGS_AS_ERRORS=1 to LocalConfig.kmk) |
d50a8773c78b10aa05c8145f2a051657d84f094a 26202 |
|
03-Feb-2010 |
vboxsync |
Broke up guest page fault and dirty page checking to avoid taking the big pgm lock. (risky change) |
c24d57d0f07ddaaa58efa612109d26fc10e43f9d 26194 |
|
03-Feb-2010 |
vboxsync |
Removed another redundant check |
a67a22561fc7a4af22a76cc49ab5a586d77f4446 26191 |
|
03-Feb-2010 |
vboxsync |
Redundant check removed |
b89a7f690e1eea15fef376a3a586472170545394 26021 |
|
25-Jan-2010 |
vboxsync |
Compile fix |
e1ab4fcc81b7783c8a04c14a3203dd72f57293b2 26018 |
|
25-Jan-2010 |
vboxsync |
Missing pending pgm sync handling in VT-x and AMD-V page fault handling. Cause of win2k3 smp guest instability and possibly many others.
Cleaned up InvalidatePage (removed obsolete code). |
e6ea1bbce4341b8288b1c9e9da56daa93feb1d08 25937 |
|
20-Jan-2010 |
vboxsync |
PGM: PGM_WITHOUT_MAPPINGS must be defined in PGMInternal.h |
6605a2c58e146cea63df460fb55c40c5707cdcfe 25935 |
|
20-Jan-2010 |
vboxsync |
PGM,CPUM: Be more careful and flexible with guest mappings on restore. (#4362) |
296e1aa2f7e2f5e33708a502923ecb74a2e75f91 25866 |
|
15-Jan-2010 |
vboxsync |
VMM: More micro optimizations. |
c878e6343c3e9807aeeb3eb35cb064ea8e83560b 25837 |
|
14-Jan-2010 |
vboxsync |
VMM: More micro optimizations around CPUM getters and predicates. |
437cb30cecc67158e35c5cc5d453cfe4eddb7dd2 25835 |
|
14-Jan-2010 |
vboxsync |
CPUM,VMM: Avoid calling CPUMGetGuestEFER until it's needed (see defect 4597). Also pushed the inlined CPUM predicates that was calling into CPUMGetSomething anyway into CPUMAllRegs.cpp since that's faster and smaller. |
ee9217e77f47a4f07c6b53e7fc0c7174570e70c4 25579 |
|
23-Dec-2009 |
vboxsync |
COmpile fix |
8c4fe3e310c213aa9601de3b116e667565e4d15f 25578 |
|
23-Dec-2009 |
vboxsync |
PGMHandlerPhysicalDeregister: don't flush the TLB as we haven't changed any PTEs yet
Trap0eHandler: only flush the TLB entry on the current cpu in case of a stale TLB entry with R/W=0; overkill to do this for all VCPUs, as we can just take faults of stale TLBs |
50d57198c1fe9ef276e4dfe9a1a79d7eab268d98 25537 |
|
21-Dec-2009 |
vboxsync |
Removed all inactive PGM pool code. |
5f14010d0da6c738c59bc9c916c2414afdb43296 25500 |
|
18-Dec-2009 |
vboxsync |
Todo added for the InvalidatePage handler |
e0341fc9c9c65532f4d494682549f309369372d4 25245 |
|
08-Dec-2009 |
vboxsync |
PGMAllShw.h: -Wshadow |
fe9672ad22c6115fce3470cabac2390e9de144dc 24978 |
|
26-Nov-2009 |
vboxsync |
Regression fix |
6cd77a666b4ea16ea457b6b487e627fe0fc775ef 24966 |
|
25-Nov-2009 |
vboxsync |
Accidental logging removed |
bf45c320837fb0d189d456646ab5c54eea72d81a 24965 |
|
25-Nov-2009 |
vboxsync |
InvalidatePage: Reset the modification counter (OpenSolaris trashes tlb entries very often) |
abcfa8f3f40121cd2178d2029b1d697523b096f9 24806 |
|
19-Nov-2009 |
vboxsync |
PGMAllBth.h: be even more careful. |
0c367a847e88805256f0ab9ac88937e6641f1149 24805 |
|
19-Nov-2009 |
vboxsync |
PGMAllBth.h: Make the page writable before remapping it RW in the dirty bit emulation. |
f0468ed96fce7ee35dffdc37d84bd702128d856f 24793 |
|
19-Nov-2009 |
vboxsync |
SSM,PGM: Added the pass number to the LiveVote callback. Improved pgmR3LiveVote a little bit. Started on real write monitored pages (#4325) - still disabled. |
1b06916966bc4aa3d33f3cfcdf42a7a6e8f618d0 24699 |
|
16-Nov-2009 |
vboxsync |
PDMAllBth.h: Wrong format specifier. |
3b64cc8531062f74735d85e0d7715b53ea511af6 24693 |
|
16-Nov-2009 |
vboxsync |
Wrong format specifier |
2d1a390af6182514573e060b2c814f45e4f0d683 24216 |
|
30-Oct-2009 |
vboxsync |
VMM: Backed out r54155 & r54154 due to XP breakage. |
0dc95ded100a03026dd70e4e865b62537bc70524 24208 |
|
30-Oct-2009 |
vboxsync |
Warning |
dd634536355579eb63fdff2ee57f7ab5b620d2f6 24207 |
|
30-Oct-2009 |
vboxsync |
Check if we need to intercept invlpg; we catch all page table updates, so there's no need unless there are dirty (unmonitored) pages.
The InvalidatePage handler should only check for dirty pages for the same reason. Restricted to R0 for now.
Again a risky change. |
2857efd62661bc77a6337a3c164bcb369b4f51ca 24041 |
|
23-Oct-2009 |
vboxsync |
Also flush dirty pages when unmapping cr3 |
dc5533b962439ceace45bdb0a5c15e39f1628f41 23485 |
|
01-Oct-2009 |
vboxsync |
PGMPAGE: Optimized tests accessing both u2HandlerPhysStateY and u2HandlerVirtStateY. (saving ~2KB on linux/amd64) |
c8074ced850b906fbc2bd446a1b148b875252030 23284 |
|
24-Sep-2009 |
vboxsync |
Correction |
acba9722fdb2a8ee387985604a89e4507bda0789 23283 |
|
24-Sep-2009 |
vboxsync |
Use atomic operations to update page table entries. |
dac32fabf3180155d9a9b09e778852e5624f7ccd 23254 |
|
23-Sep-2009 |
vboxsync |
Empty line |
f1c295317cd1c7ebf197445bc636a32cbaa4b9a8 23252 |
|
23-Sep-2009 |
vboxsync |
Backed out 52727; was correct |
2a9a5f63186987ff2f089413c65aab2729c1681e 23251 |
|
23-Sep-2009 |
vboxsync |
Used the wrong macro |
b66e14ead868d803cc49957974c34c4ea47346b4 23250 |
|
23-Sep-2009 |
vboxsync |
Deal with dirty page tracking and monitored pages. |
4adf8d9f73a46196eee17b66fdaa0d100a76dc43 23116 |
|
18-Sep-2009 |
vboxsync |
Noisy logging |
b0500bf4ec1b0cd231c6651ff2f5337e01ca93b1 23097 |
|
17-Sep-2009 |
vboxsync |
Distinguish between single page sync operations due to non-present or write faults. |
e17f587595bd5d3a7be56a892e3fd3a0ef83d268 22890 |
|
10-Sep-2009 |
vboxsync |
VM::cCPUs -> VM::cCpus so it matches all the other cCpus and aCpus members. |
2b5a4cf3d77ab62dcbd882115b6d497547b20d29 22600 |
|
31-Aug-2009 |
vboxsync |
Removed unnecessary EPT invlpg calls.
Invalidate the page who's R/W attribute was changed. |
fdc5224bd8d9a60af82da5809e3d6729c9bc69cb 22473 |
|
26-Aug-2009 |
vboxsync |
Disabled experiment with optimized dirty PTs. |
0346214b50356346da480fb73f7d19b66a4e9e6c 22327 |
|
18-Aug-2009 |
vboxsync |
Disabled rarely useful statistics as they pollute the log |
b0d2ede40a3ebf93f7793af9ce9e5c913cdcdbba 22326 |
|
18-Aug-2009 |
vboxsync |
Disabled rarely useful statistics as they pollute the log |
d6ee54769ba74a3799a2f4511aba0fbc9f276ffd 22316 |
|
18-Aug-2009 |
vboxsync |
WP checking is not relevant for nested paging & EPT. |
aadffbdb31f7f529dc051a42d496c65f69847456 22169 |
|
11-Aug-2009 |
vboxsync |
Avoid incorrect reuse of big pages: take the NX bit into account too. |
7be55816e78ba5b3f590e52bcb192c7894717af4 21966 |
|
04-Aug-2009 |
vboxsync |
Workaround for fatal cached page table free. |
3f6d9bc043b47558ea407f9cd618f0dd177b3864 21175 |
|
02-Jul-2009 |
vboxsync |
Don't release the PGM lock when calling the PGM pool handler. (pointless as we'll request the lock again immediately) |
714b11dcd13ea2bcddbc1f9e25aa412cb405ff88 21123 |
|
01-Jul-2009 |
vboxsync |
Removed redundant check |
b039aa4c852d9d1b19fbb281f206e693fa4fda05 21116 |
|
01-Jul-2009 |
vboxsync |
Adjusted assertion |
9982c9e62bb671f5f2c6d4a954b40647adf56031 21115 |
|
01-Jul-2009 |
vboxsync |
Moved CR0.WP=0 handling up a bit |
18e02d8d60d50fb098bbf009186f36bfb430a780 20871 |
|
24-Jun-2009 |
vboxsync |
VMM: Make sure there is enough room for a few physical handler notification before we disable ring-3 calls. Partial VMM[GC|R0]CallHost unification. |
b9789db9e2587212305fc440e3ec40599559a6e7 20795 |
|
22-Jun-2009 |
vboxsync |
VMM: Fixed guest PAE issues on the mac. |
aecdea8745180791ba6060f8b8f7672b6e6f24e8 20783 |
|
22-Jun-2009 |
vboxsync |
Less restrictive assertion for smp guests. |
dcc17b8e0f81054c120fa35fc6b68ff65ccf5563 20772 |
|
22-Jun-2009 |
vboxsync |
More safety |
cb06e3f411d6ced5515682cddfbfa8ec463c6ff3 20762 |
|
22-Jun-2009 |
vboxsync |
More PGM locking |
c60ca5235c1a8b3c4f348e75167e93417b05f7c4 20684 |
|
18-Jun-2009 |
vboxsync |
PGMAllBth.h: shut up gcc warnings. |
ffea7b3614d7843b53411a7636b66195b52d524d 20671 |
|
17-Jun-2009 |
vboxsync |
Bigger lock for the pagefault handler.
Avoid deadlocks when syncing notification handlers with our recompiler. |
3cbb4f9a6a320e58ed398ef7aaa004cc8727abc5 20530 |
|
13-Jun-2009 |
vboxsync |
VMM: remove DISCPUSTATE from the stack. |
99b52f2596317deb5589eec524285922a97fe636 20519 |
|
12-Jun-2009 |
vboxsync |
Disabled bogus assertion in nested paging mode |
013bdbd0cd83ebfba45641fa2cd57ec6e97eb396 20518 |
|
12-Jun-2009 |
vboxsync |
Assertion message |
ddfb39e2f8eea6e726aed5b0c3f9a8613e7b8d6d 20465 |
|
10-Jun-2009 |
vboxsync |
Logging fix |
590bfe12ce22cd3716448fbb9f4dc51664bfe5e2 20374 |
|
08-Jun-2009 |
vboxsync |
*: s/RT_\(BEGIN|END\)_DECLS/RT_C_DECLS_\1/g |
2af0dd14e6c1c9297537694bd8ff24ae5838620c 20365 |
|
08-Jun-2009 |
vboxsync |
VMM: __BEGIN_DECLS -> RT_BEGIN_DECLS; __END_DECLS -> RT_END_DECLS. |
e862b2af80d4e75ed789c6aac07b9566c983aeb3 20228 |
|
03-Jun-2009 |
vboxsync |
Fixed VT-x state loading failure. |
74991d7531692858fd22acf371a7ee941567977c 20135 |
|
29-May-2009 |
vboxsync |
Prevent reuse of cached larged pages with different access attributes. |
a23ceb1bffe284d92ab25e6ea3da75172aad878b 20129 |
|
28-May-2009 |
vboxsync |
Backed out 47892-5 |
2bb24ea1aa00aa5689beb1f940775fe735301ea4 20128 |
|
28-May-2009 |
vboxsync |
Bug fix |
e64eabfa44af470c540c28794b9d1a26c9f2e676 20127 |
|
28-May-2009 |
vboxsync |
Cleaned up |
9da1bf65ef30a560c65a0ec4dffb75ea75bb2be1 20126 |
|
28-May-2009 |
vboxsync |
Wrong place |
953b272ae1f08ea6fda3fb9d618dcb77b795916f 20125 |
|
28-May-2009 |
vboxsync |
Prevent reuse of cached large pages with different access attributes. |
bc11e048fcbb291722c30254b82c4679ab8990ea 20069 |
|
27-May-2009 |
vboxsync |
Loosened assertion for the guest SMP case. |
757023c8d581a612de4d228a17d2f686e5129f30 20068 |
|
27-May-2009 |
vboxsync |
Trying to get rid of annoying assertions |
11b04d5e9a0e449814ac17dd605920b4e5b4cf9d 20065 |
|
27-May-2009 |
vboxsync |
Logging |
b80218788444df820ebd9daa890de20ffce1e23e 20064 |
|
27-May-2009 |
vboxsync |
SyncPage: deal with missing PDE; can be changed behind our back by another VCPU |
b5a890ec71a7a620319fc0c97a570b5d9eb237a4 20061 |
|
27-May-2009 |
vboxsync |
Made assertion less strict in SMP mode |
2c3f1291a717b4c4943302e9b24c077782d10355 19871 |
|
20-May-2009 |
vboxsync |
Protect SyncPage with a lock as well. |
54a06187e74d85d7023d43f8a47e8c3972e6d0f5 19835 |
|
19-May-2009 |
vboxsync |
Comment update |
688b55d5387d8b7c508d30141fe1a96eadecb988 19833 |
|
19-May-2009 |
vboxsync |
PGM_INVL_GUEST_TLBS -> PGM_INVL_VCPU_TLBS |
611d60e864ddb9c200e6a69f402b25e8650c601c 19808 |
|
19-May-2009 |
vboxsync |
Invalidate is per VCPU. |
1b905f7766005546c0ab6799727d513ff4829335 19791 |
|
18-May-2009 |
vboxsync |
Added remark to InvalidatePage about TLB shootdowns. |
ee4859ccbf53d481071dac709adcad1ddf57a395 19790 |
|
18-May-2009 |
vboxsync |
Protect InvalidatePage with the pgm lock. |
d011e03a419b3cfbc30fabc1b18cc3ba3774f013 19789 |
|
18-May-2009 |
vboxsync |
Atomic write |
45998f6120a1fac488283061ca862ae9c2cfb212 19788 |
|
18-May-2009 |
vboxsync |
Restrict some invlpg to RC only. |
6fcd20f3616946a878f838a5b86dbb9bdc1cf032 19785 |
|
18-May-2009 |
vboxsync |
Locking should not be required after all |
dd7a84f153b54bd6d10d487911b9ef1bf34c5d54 19784 |
|
18-May-2009 |
vboxsync |
Use PDM lock for protecting pdm queue management. |
2f9264b639f46225de66e46619693264fa988602 19782 |
|
18-May-2009 |
vboxsync |
Missing unlocks |
a637e88400c9ede78b155d46625a4081f885de39 19781 |
|
18-May-2009 |
vboxsync |
PrefetchPage: Locking |
1475f10713478f11503ff52947c07a7ee6b7031d 19779 |
|
18-May-2009 |
vboxsync |
Only need stale TLB detection in ring 0. |
0dfb7b28f32f468fda2e189c973d4902d6ab6dad 19778 |
|
18-May-2009 |
vboxsync |
Fixed the stale TLB detection |
a7d237d8103fe4926a308025e14e97c165566570 19777 |
|
18-May-2009 |
vboxsync |
Atomic write |
329317fd7d7cb0d4f2b127ab5e49001f34648349 19776 |
|
18-May-2009 |
vboxsync |
Reapplied fixed 47403 |
c9ce78aecb99bded4bae6c303bb35eac311ffc42 19775 |
|
18-May-2009 |
vboxsync |
Reapplied 47403 |
50269b2a88fa5fa9e49183b22055b6aee9992d2c 19772 |
|
18-May-2009 |
vboxsync |
Backed out 47401-3 for now |
cd7f63c06488ad037c095404e7baed621da75511 19771 |
|
18-May-2009 |
vboxsync |
More atomic writes |
a4e43553bf5aa819edc68e05abf5086f2d770b10 19770 |
|
18-May-2009 |
vboxsync |
Missing unlocks in error cases |
3bb39c60ac4d0de8e31048af759eaca6208e4443 19769 |
|
18-May-2009 |
vboxsync |
Deal with stale TLB entries for harmless page entry updates.
Wider locking. |
713017d50d9158c0d745d9872f9e75f5c04ba18c 19765 |
|
18-May-2009 |
vboxsync |
Atomically update page table/directory entries. |
567c83ddf1f21fe3cdd571042dfec7f84e2843aa 19691 |
|
14-May-2009 |
vboxsync |
SyncPT must be protected by a lock. |
41b8e304f190c3d901bcfee5e73c11bc2c42ce13 19329 |
|
04-May-2009 |
vboxsync |
Removed pgmPoolGetPageByHCPhys. |
e9304f87bb1ea09de6efd1f2d8245d61c99f7af9 19195 |
|
27-Apr-2009 |
vboxsync |
Fixed possible corruption due to RC cache entry reuse. |
e0b9d3c357adf9b7d05f55540e86f22943fc4b23 19141 |
|
23-Apr-2009 |
vboxsync |
Action flags breakup.
Fixed PGM saved state loading of 2.2.2 images.
Reduced hacks in PATM state loading (fixups). |
ebeba5ff831e671655fb82b4287aadaa0f5367f1 19015 |
|
20-Apr-2009 |
vboxsync |
Split up TRPM. (guest SMP) |
d45f7f7fe0c28b500b45b2dc88d7a04f4c0be6b8 18992 |
|
17-Apr-2009 |
vboxsync |
More PGM api changes |
87fe4d73d7e6e53fbcec40dc6be2372479851cd4 18988 |
|
17-Apr-2009 |
vboxsync |
PGM api changes |
3cf963d8e186ef935468e4425a522ae787d93e14 18947 |
|
16-Apr-2009 |
vboxsync |
Mac compiler complains a lot |
42c1972c22e09797b4b24afbd0ec114ed076c37c 18927 |
|
16-Apr-2009 |
vboxsync |
Big step to separate VMM data structures for guest SMP. (pgm, em) |
8244b8f56781b7421c6b42261e42baf3d0b4c64b 18835 |
|
07-Apr-2009 |
vboxsync |
PGMAllBth.h: Fix for CR3 mismatch assertion in strict builds when nipping back to ring-3 to grab the PGM lock. Grab the lock before entering dangerous waters... |
29007d2297c03e015a08607be26338bf2dab00df 18790 |
|
06-Apr-2009 |
vboxsync |
PGMAllBth.h: todos. |
0d0687880459ea33aa3e9c5a74d736edcb0937de 18666 |
|
03-Apr-2009 |
vboxsync |
VMM: Clean out the VBOX_WITH_NEW_PHYS_CODE #ifdefs. (part 2) |
ff78b877ed7acd25e2d384570a938441455d6a95 18617 |
|
02-Apr-2009 |
vboxsync |
PGM,EM: Handle out of memory situations more gracefully - part 1. New debugger commands: .pgmerror and .pgmerroroff. |
3c7cee28a97675a7206fd99445588a7f0b69b9f9 18330 |
|
26-Mar-2009 |
vboxsync |
MapCR3: return VERR_INTERNAL_ERROR_2. |
64d004095ebf55be2e68fee90e0d7fca9706a874 18192 |
|
24-Mar-2009 |
vboxsync |
nc |
485b22f9a4dd2c128a6c85f1164a8de93db62af2 18125 |
|
22-Mar-2009 |
vboxsync |
PGM: MapCR3 anywhere again (new phys code). |
34aeec2d93faf945f74ba0ba051eaecb87870ad5 18093 |
|
19-Mar-2009 |
vboxsync |
More verbose assertion |
750d4d0506a38b2e80c997075d40aad474e675fb 17667 |
|
11-Mar-2009 |
vboxsync |
pgmPoolAlloc no longer fails with non-fatal errors. |
b009f0780a29fa1d29a9b470c9b2127010df2872 17660 |
|
11-Mar-2009 |
vboxsync |
PGM: Saved state hacking, fA20Enabled is now bool, shadow ROM bug fix, and mapping chunk limit correction for 32-bit hosts. |
f313ddae4a5e83adf2cd2e3d9ba2ae6fefc448d4 17644 |
|
10-Mar-2009 |
vboxsync |
Fixed one lock |
bd59ae9e043954cbd80f978c8369fdb15f2c404b 17643 |
|
10-Mar-2009 |
vboxsync |
Some more locking. We really need a better solution. |
9dd5680d0a9f78eac6a907e4fa197272699f417b 17641 |
|
10-Mar-2009 |
vboxsync |
Exclude some unused code in RC. |
667009dadf5473101ffec18ddc317230abdc50e6 17616 |
|
10-Mar-2009 |
vboxsync |
Inform SELM about shadow CR3 changes. |
46b1ac1c3302722f93e787d0f0693965e88f5cee 17586 |
|
09-Mar-2009 |
vboxsync |
Removed all dead non-VBOX_WITH_PGMPOOL_PAGING_ONLY code. |
1ce415c61b519fd008de8a50dcc386720bac2efc 17562 |
|
09-Mar-2009 |
vboxsync |
Backed out 44000; code was already inactive |
88c9a875d98ac61b3d468b832e566c118e1a4838 17561 |
|
09-Mar-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: doesn't apply here |
11c599b71a8c202a506b8623af57432c3d415dda 17509 |
|
07-Mar-2009 |
vboxsync |
PGM: Moved the page pool PT flushing code in the access handler bits to where it belongs and called it pgmPoolTrackFlushGCPhys. Fixed a status code corruption bug in PGMR3PhysTlbGCPhys2Ptr (new phys). Made lazy zero page replacement code work in the new code, it's disabled by default because it frequently requires flushing the shadow page pool because the tracking code assuming the HCPhys of a PGMPAGE is unique and never shared. |
43376d33564c03a42b5af38c17ea8287fc1bec7b 17505 |
|
06-Mar-2009 |
vboxsync |
PGM: MapCR3 hack for the new code, fixing PGMFlushTLB status propagation error in emUpdateCRx. |
e27645d13c6aba33ceea5b5b515d2e54b0dd5a0f 17486 |
|
06-Mar-2009 |
vboxsync |
Updates |
410f9d3ddfa779d92ca3662eb68fea60d314199a 17483 |
|
06-Mar-2009 |
vboxsync |
Cleaned up nested paging init and removed obsolete structure members. |
7328d8bec2798b4b8f02b3c7fc4f27d5661dc84f 17438 |
|
06-Mar-2009 |
vboxsync |
PGM: Some REM hacking for the ROM registration and some in-progress page sync hacking... |
ccd4cc1060542fab8d8d48505adc07af85d8aba0 17432 |
|
06-Mar-2009 |
vboxsync |
PGM,GMM: Filling in missing bits and fixing some bugs. |
3518b2ce0d5a4e029b96fdb6336894b49db10a18 17422 |
|
05-Mar-2009 |
vboxsync |
VMM: Do not forcibly enable RC logging in VMMGCLogEnabled! Renamed it to VMMGCLogRestore and made VMMGCLogDisable return a state indicator that should be passed to it. |
07ab3ea8c820f7fbfb587bdff34b4fd31d586b94 17398 |
|
05-Mar-2009 |
vboxsync |
Adjusted assertion again |
8a6fc1da1684d5e3ad9144cec1cfca3211fa05ce 17395 |
|
05-Mar-2009 |
vboxsync |
Adjusted assertion |
88f323f0756f302b5fa556549767897bbf935bb8 17372 |
|
05-Mar-2009 |
vboxsync |
PGM,MM: Made VBOX_WITH_NEW_PHYS_CODE compile and link. |
bee80f3f77b15c166b7debd080a70e5231ba6751 17316 |
|
03-Mar-2009 |
vboxsync |
PGM: Use %R[pgmpage] instead of logging PGMPAGE::HCPhys. |
fa209a078dd0f7766806071b3b0fdc614708c407 17294 |
|
03-Mar-2009 |
vboxsync |
PGM: started looking at direct access of PGMPAGE::HCPhys and the HCPtr/mapping nightmare. |
943eda4b0dfb8421c1c2fca62d354cc65ec9068d 17290 |
|
03-Mar-2009 |
vboxsync |
PGM: -MM_RAM_FLAGS_IDX_SHIFT, -MM_RAM_FLAGS_IDX_MASK |
7a066e597afbbd02a36b2d7af4cbfd06ec32d259 17289 |
|
03-Mar-2009 |
vboxsync |
PGMAllBth.h: Fixed wrong HCPhys mask in EPT mode (EPT_PTE_PG_MASK=X86_PTE_PAE_PG_MASK_FULL, while we must use X86_PTE_PAE_PG_MASK against PGMPAGE::HCPhys, which omits bits 48 thru 51). Could in theory cause these bits to be wrong, although I'm not so sure intel cares. |
c65c00499651bf2c4f270edd76c7f30aa30e85aa 17288 |
|
03-Mar-2009 |
vboxsync |
PGM: Some MM_RAM_FLAGS_IDX_SHIFT cleanup. |
65ff326ac3915638f5c4e79c2c3e836aeeeb86de 17285 |
|
03-Mar-2009 |
vboxsync |
PGM: Gone are MM_RAM_FLAGS_CREFS_SHIFT and MM_RAM_FLAGS_CREFS_MASK. |
57399ab65e2825c324fb9dcb4642d4ae2c232509 17279 |
|
03-Mar-2009 |
vboxsync |
PGM,MM: Attacking the shadow page pool tracking info stored in PGMPPAGE, replacing direct access with PGM_PAGE_SET/GET_XXXX access. Required to get rid of MM_RAM_FLAGS_* and to be able to restructure PGMPAGE. MM_RAM_FLAGS_NO_REFS_MASK fully eliminated already. |
dfddc59b40a5ae34361b34653bf3d71e420a85f2 17216 |
|
27-Feb-2009 |
vboxsync |
Removed unnecessary invlpg for HWVirtEx mode |
56987583fa505206fcb803cef7f0caad9027c1ce 17215 |
|
27-Feb-2009 |
vboxsync |
Split up the definitions and the guest code. Otherwise we'll end up using e.g. wrong masks in Bth code. |
91d834b667d6d8a2182a0adfd17aaab3dfbc1b13 17207 |
|
27-Feb-2009 |
vboxsync |
Corrected assertion for nested paging |
5275792d4c9845ec66b3db0339356a7323aa523c 17206 |
|
27-Feb-2009 |
vboxsync |
Compile fix |
3be806557b77686cd6976d7545fba731fdd7149e 17146 |
|
26-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: Disable logging during a critical section in MapCR3 (raw mode only). |
252acd1a1fd57f76466baa837f99a0820e6b6f36 17142 |
|
25-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: fixed saved state restore |
44ae8e78ba212d0ec666d147b5f083e91b5a5b4a 17137 |
|
25-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: explicit locking of root CR3 pages |
5e0e35e65847aae0955835d5c4b4e76143e580c1 17135 |
|
25-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: Disallow ring 3 jumps when in an inconsistent state during CR3 switching. |
9400b66fe4820424797e495f570f36b698333167 17134 |
|
25-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: updates |
97d7e01c4d2e10e7ac2e086d7d3a7b75cb4b108b 17059 |
|
24-Feb-2009 |
vboxsync |
Updates |
bfcd0b11ad299b9bbb9f99be18e5e336d2a86dc6 17058 |
|
24-Feb-2009 |
vboxsync |
More updates for temp locks. |
16792b73c78f101caf2d512d4d41abaf5d4be43c 17001 |
|
23-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: update |
a6c36cd330fbbe3d5019d79be593227e180aa47c 16917 |
|
18-Feb-2009 |
vboxsync |
Inverted assertion check |
02b73dfb723f5b242039306b12d4324f2426420a 16902 |
|
18-Feb-2009 |
vboxsync |
Assertion added |
32b28388143e8c86a1b599e155ef22a739b8dd7e 16901 |
|
18-Feb-2009 |
vboxsync |
Some comments added |
3f6d99d3bcd6189ee444c87c6b7a18013b2bebec 16898 |
|
18-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: simplified SyncCR3; introduced PGMMapResolveConflicts |
8257b04810d609a9b3b30bb6a43765339894eb19 16866 |
|
17-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: rc fix |
4a296be15f381ac7f3506e4eb2861627d062fee3 16859 |
|
17-Feb-2009 |
vboxsync |
Load hypervisor CR3 from CPUM (instead of hardcoded fixups in the switchers). Dangerous change. Watch for regressions. |
2edea3c3d2f9b2afd1fcc8ac4e045df2c745b425 16842 |
|
17-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: update cr3 in RC |
4479d221e12b043711464af981ef0f799c374496 16840 |
|
17-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: updates for raw mode |
cef739c6822f5cc74f67f6b7218487b6e4f59a6e 16800 |
|
16-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: SyncCR3 fixes |
858595de1775e2f065504770c514814ec1acdb40 16798 |
|
16-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: deactivate hypervisor mappings in old CR3 root |
616a0c0896edec0856c44698196cd82f0ded406b 16792 |
|
16-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: updates for raw mode |
2c080f6fc6ab70995d6390eeaa199d590c1a34d5 16780 |
|
16-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: typo |
ffa52b282225b48c4656480f389cd9c3bea71cb2 16659 |
|
11-Feb-2009 |
vboxsync |
More logging |
75277affcc9892d09548b5b9c6359cda0fda9a61 16658 |
|
11-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: update |
451885806845f32b94d7bd266f490e6e7060b535 16636 |
|
10-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: updates |
58ea955d6b9cb05f52ced316fadfe16e97f8b025 16593 |
|
09-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: inverted check |
39f76759232d588a22982f64f383c4e20366d6c0 16592 |
|
09-Feb-2009 |
vboxsync |
Setting the accessed bit of PAE PDPTs causes an VT-x invalid guest state exit. |
6aa2270fe6daca24368a6d8a1d8a139231d00cb2 16591 |
|
09-Feb-2009 |
vboxsync |
Same here |
2daf0e598c7675ad40bfac76d53b6616bdbebf42 16590 |
|
09-Feb-2009 |
vboxsync |
rw/us are reserved for PAE pdpte's. |
6a99f6112f9f7e78794e3d76d40737aa3983f4aa 16582 |
|
09-Feb-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: updates |
befb58e44ac456581535a5581a8fb6b87f5e3a52 16424 |
|
30-Jan-2009 |
vboxsync |
VBOX_WITH_PGMPOOL_PAGING_ONLY: fixed PrefetchPage for PAE shadow paging |
1e3576bc616045a5585190fcbc221648d17b23af 16419 |
|
30-Jan-2009 |
vboxsync |
Store the iUser & iUserTable of the root shadow page table in PGM. |
756d197dbc3b884625fa54686bad0ceec9f94ea3 16407 |
|
30-Jan-2009 |
vboxsync |
More logging |
61a20f87ecfcf850a70f6832abfb56e581bc6593 16376 |
|
29-Jan-2009 |
vboxsync |
Paging updates. Deal with mode switches and cr3 updates. |
1426cc093ab6f03dc7481677cc54995bf7c06461 16321 |
|
28-Jan-2009 |
vboxsync |
More paging updates |
5db1d52ffbcaa46c3d944c6c2d9c552306817d9a 16317 |
|
28-Jan-2009 |
vboxsync |
Moved Map- and UnmapCR3 to Bth as they affect both guest and shadow structures. |
bd7bffdb1ee3b976bbb671e70840bca2c08504d8 16260 |
|
27-Jan-2009 |
vboxsync |
More paging updates |
86b687e7808a36be33c43ae58adc8ab22d378fea 16232 |
|
26-Jan-2009 |
vboxsync |
More shadow paging updates (disabled) |
3080f6c0871099df43a4e91b31894d9c2b1369a8 16203 |
|
23-Jan-2009 |
vboxsync |
Updates in preparation for PGM pool based paging everywhere. |
bacaf5667ff6b292c78603ed3a6f8ef0fb02e400 16177 |
|
22-Jan-2009 |
vboxsync |
Another Mac compile fix |
b9acb8b1d78c488d1498f7568284f4ed0c689851 16172 |
|
22-Jan-2009 |
vboxsync |
Moved amd64 paging data to unified shadow paging section. |
26824086a3f6b36cd911058f1d9b4c0b944706fb 15410 |
|
13-Dec-2008 |
vboxsync |
PGM: Deal with pgmPoolAlloc failure in MapCR3 without relying on having to clear the pool. The MapCR3 action will be postponed to SyncCR3. |
a3096f27bbd73b32a7a8b380f0950cc78350f005 15228 |
|
10-Dec-2008 |
vboxsync |
backed out 40461. (PGMAllBth.h) |
a037a5c9b8d88b719edaefc0b755d290cd865f40 15227 |
|
10-Dec-2008 |
vboxsync |
PGMAllBth: Fixed typo in CalcKind (harmless), did some predicting and cleanup in SyncCR3 (PAE/32-bit). |
6aa8adb8cae7a6385e9f0e50190bd59beef4b8a4 14868 |
|
01-Dec-2008 |
vboxsync |
PGMR0DynMap: some predictions based on statistics, 5 more stats. |
5aa784a1ddb77b33952a2ab0b2707d84d48fcb7e 14755 |
|
28-Nov-2008 |
vboxsync |
#1865: Converted 4 PGM*2HC* conversion functions to RTR3PTR. |
8befce7403eb5ad1307761cfcfb65e5845e4e73f 14681 |
|
27-Nov-2008 |
vboxsync |
PGMAllBth.h: assertion text. |
b76dcff6c9d35ade3155039689bfc30057481330 14677 |
|
27-Nov-2008 |
vboxsync |
SyncCR3: Fixed pPDEDst walking with GST=32/SHW=PAE. Fixed assertion. |
4bb8b623093edfe938b5b3d99f2197395eed5f35 14392 |
|
20-Nov-2008 |
vboxsync |
In PAE shadow mode we must always reactivate PDPT entries. (Solaris 10U6 reboot) |
45615a5417d988abf34f831591c905cc3db3d2ab 14364 |
|
19-Nov-2008 |
vboxsync |
Preparations |
c01526c140e6e82fd08bd5d5adfb0774ba2b6690 14154 |
|
13-Nov-2008 |
vboxsync |
PGM: pGuestPD* -> pGst32BitPd* |
6357f688d3ef2cebb368c3c840731663a595d354 14153 |
|
13-Nov-2008 |
vboxsync |
PGM: pGstPaePDPT -> pGstPaePdpt |
3cfa84e65f00d41b18366c42671e5c1d596c5197 14147 |
|
13-Nov-2008 |
vboxsync |
#1865: PGM - 32-bit shadown PD pointer converted. |
3cf3eede281c1aa32944353bb2726bdda3b9c829 14138 |
|
12-Nov-2008 |
vboxsync |
build fix |
635e435d281bb83c0c544ad6572d4151c7c73278 14133 |
|
12-Nov-2008 |
vboxsync |
#1865: final pae change. |
ca18c35cbebca7011fb9d0958c08a54644fe580f 14131 |
|
12-Nov-2008 |
vboxsync |
PGM: Prevent fixated mappings from crossing a PD boundrary (guest+shadow PAE). |
7a5fb4558a3a8888597fb0311cf6d7f6f24ea0fe 14103 |
|
11-Nov-2008 |
vboxsync |
#1865: shadow pae, the seventh wave. |
f8bd91b2e1a9a0a1705b786375014b8012a0e3a5 14102 |
|
11-Nov-2008 |
vboxsync |
#1865: shadow pae, the sixth bit. |
1e603e041b22e3bb993fd6683882e1d3ad00ae4c 14101 |
|
11-Nov-2008 |
vboxsync |
Use pgmPoolFree. |
ce832d5c94bdbe1eacc83663b261bd18048fe108 14099 |
|
11-Nov-2008 |
vboxsync |
#1865: shadow pae, the fifth bit. (removed dead code) |
4ed13da9f6cb81fdd3db3524c7ed79f421e16032 14096 |
|
11-Nov-2008 |
vboxsync |
#1865: shadow pae, the forth bit. (dead code checks) |
0e9168bb25c5b54fa3cc110219c7b53726f0a957 14095 |
|
11-Nov-2008 |
vboxsync |
#1865: shadow pae, the third bit. |
14d63a79cba5f74799d8fd380927f5dc87bc0bdc 14094 |
|
11-Nov-2008 |
vboxsync |
#1865: shadow pae, the second bit. |
57ad8aa3a9c059225e996dcf315942c2e8628f17 14093 |
|
11-Nov-2008 |
vboxsync |
#1865: shadow pae, the first bit. |
148514a0d25c3bc8866749f4b6994a6f66b4adfe 14038 |
|
10-Nov-2008 |
vboxsync |
#1865: PGM - and another one. |
49bd9c2e1effcb74e0535b5f8d65b308d72831d4 14010 |
|
10-Nov-2008 |
vboxsync |
#1865: PGM - one more down. |
f2e5ce7cd98acbb39c6497af057ec3e1a50058e4 13991 |
|
10-Nov-2008 |
vboxsync |
#1865: More PGM cleanup. |
3cafd14c52c6da58524065dad928a4d2593e5157 13937 |
|
06-Nov-2008 |
vboxsync |
PGM: RTGCUINTPTR -> RTGCPTR. |
5e96135cb5356e42ae8ecebd912c54cda06791d0 13936 |
|
06-Nov-2008 |
vboxsync |
PGM: removed a bunch of unnecessary RTGCUINTPTR/RTGCPTR casts. (RTGCPTR used to be void * in RC, now RTGCUINTPTR == RTGCPTR.) |
a507a9ef736e6075c1a15bbf9af3516dd4744648 13935 |
|
06-Nov-2008 |
vboxsync |
PGM: One more CTXSUFF (guest AMD64 root). |
754341be40c49685913ed219ce8bd4f49068d837 13933 |
|
06-Nov-2008 |
vboxsync |
PGM: PAE guest mappings, assert GCPtr being 32-bit. |
eefe1f58ec0c7c9bb1e90be2a22cae7b58099606 13919 |
|
06-Nov-2008 |
vboxsync |
VMM: 3 R3R0PTRTYPE members down (a bunch left to go). |
7e960d3a0a8a3a84d7aba2cca45d72b1c31cc97b 13832 |
|
05-Nov-2008 |
vboxsync |
IN_GC -> IN_RC. |
09fb3a10eb3d58047973f0652f546d1f2ded2068 13826 |
|
05-Nov-2008 |
vboxsync |
VMM: And a couple of left overs. phew. |
4e9f150666f4a6080824a98ee0c5881f39b1177c 13825 |
|
05-Nov-2008 |
vboxsync |
VMM: %VX64 -> %RX64 |
b1c3cdef473df2fbc621d5da81acc82dbfb8a11a 13824 |
|
05-Nov-2008 |
vboxsync |
VMM: %VGp -> %RGp |
08bc90fc2848c80bf8270bedc883745b8398e186 13823 |
|
05-Nov-2008 |
vboxsync |
VMM: %VGv -> %RGv |
79b07d7ae28b916539c6397e010a5cc05e300551 13820 |
|
05-Nov-2008 |
vboxsync |
VMM: %VG* inspection - an awfaul lot of these, hope I got it all right... |
8da94e4d6813b682cdb38eb9b151ada72d91b40e 13819 |
|
05-Nov-2008 |
vboxsync |
VMM: %VH* -> %RH*. |
b3547e42ce5221377866e3fa041b3086b7cc1562 13818 |
|
04-Nov-2008 |
vboxsync |
VMM: %Vrc -> %Rrc, %Vra -> %Rra. |
1a25adaca81841abf5e6cdfed02eaff64941357d 13816 |
|
04-Nov-2008 |
vboxsync |
VMM: VBOX_SUCCESS -> RT_SUCCESS, VBOX_FAILURE -> RT_FAILURE. |
7378bb063a588194a86a2241be539286a838f48d 13398 |
|
20-Oct-2008 |
vboxsync |
Don't check disabled physical access handlers |
18f52d8d1134c20b3d632c43d55bd84dd022ae57 13397 |
|
20-Oct-2008 |
vboxsync |
Backed out 38140 |
f41d57a1f06aa3089c1ae7af32cc03f899ff7730 13396 |
|
20-Oct-2008 |
vboxsync |
Correction |
35bc30820cb5ec2bcdbbeb75408e0ba68010323d 13395 |
|
20-Oct-2008 |
vboxsync |
Corrected assertion |
34bc80786c6507b1781a6dc0f336f23b3f284c4f 13394 |
|
20-Oct-2008 |
vboxsync |
Implement PrefetchPage for EPT when in real or protected mode without paging. |
ea100b7dd58c77505cd43fd7025b129041c860d8 13235 |
|
13-Oct-2008 |
vboxsync |
PGM: Merged PGMGCInvalidatePage into PGMInvalidatePage nad fixed the callers of it to handle the return codes correctly. other cleanup. |
407c12cfc00a01e282ea5ddab828c4d5d234f015 13232 |
|
13-Oct-2008 |
vboxsync |
#1865: changed some validation into pure assertions or assert-return. added some todos, update a comment or five and adjusted lots of whitespace. |
f2664e7c803a80455354f6de4c25f09949709577 13203 |
|
13-Oct-2008 |
vboxsync |
Missing shadow cases for EPT. |
f9b15f4c9552687388aecaaff11cbb3172679bc5 13188 |
|
11-Oct-2008 |
vboxsync |
PGM: polish. |
a66e51cd9a022e4b7fc63a033ca157794d5db349 13134 |
|
09-Oct-2008 |
vboxsync |
EPT: set the ignore PAT and WB cache flags in page table entries. |
fe46b31a4512cec77977c6356aa97d1ab165de24 13126 |
|
09-Oct-2008 |
vboxsync |
Set u1Execute too. |
622d9acdd38e3c53fe4fc028469b5147877fb306 13122 |
|
09-Oct-2008 |
vboxsync |
Corrected return value checks for PGMShwGetEPTPDPtr. |
3284dfc4979194ee7d696e235cdc6f67c79dd6a9 13121 |
|
09-Oct-2008 |
vboxsync |
Wrong mask |
0e881ae282002c3df319fb8a8e2b264ea0a324f1 13098 |
|
08-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
fe6477f53cf20d55ac0c197fe8f0b8ca9c68a08e 13087 |
|
08-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
a9981806c72edadef6cccd253f4747c35677e9a1 13085 |
|
08-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
1d94b54d4bd24729c7383e89aa95518204192b22 13067 |
|
08-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
15d1a4fb2793e0a0fc93f6da613c786bcec0355a 13063 |
|
08-Oct-2008 |
vboxsync |
Corrected return value checks for PGMShwGetLongModePDPtr & PGMShwSyncLongModePDPtr. |
6b022885f2cb6a55167609edecd89570cd80001d 13062 |
|
08-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
50df3da42ff6589b0ecc4f50f2288811bc370186 13060 |
|
08-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
6ed48beeb7f2ef925ae5bd091e1e475eae18f939 13046 |
|
07-Oct-2008 |
vboxsync |
#1865: Eliminated two duplicate members of PGMVIRTHANDLER while at it. |
a54b39668ee07bfc7d30ff54243b9ee423cfd917 13045 |
|
07-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
09d4e754a010196fe0d9d972b6ccc005ecb8b116 13042 |
|
07-Oct-2008 |
vboxsync |
#1865: More PGM changes. |
6f516ad9911d9037a18778742caa955fe362f8ff 13035 |
|
07-Oct-2008 |
vboxsync |
#1865: More PGM... |
897680ad5055effa568ebf266b3feac8be034642 13019 |
|
06-Oct-2008 |
vboxsync |
#1865: PGM ... |
fa5e201ccb10ee6001d8d4afe9782a0a81a20051 12999 |
|
06-Oct-2008 |
vboxsync |
Same applies to return code checks for PGMShwGetEPTPDPtr. |
3726ae68d9d25bf595498cc74bd1da9531741f73 12997 |
|
06-Oct-2008 |
vboxsync |
Fixed regressions in return code checking |
3600641b05af6ab695b4623ed40c2b7bd1d19dad 12994 |
|
06-Oct-2008 |
vboxsync |
Cleanup |
9ad5e3912962c3dbccc1afc4e7d62890fe906814 12989 |
|
06-Oct-2008 |
vboxsync |
VMM + VBox/cdefs.h: consolidated all the XYZ*DECLS of the VMM into VMM*DECL. Removed dead DECL and IN_XYZ* macros. |
6d299e273a1b143cbe6e66cc7810f73ba0cd9488 12964 |
|
03-Oct-2008 |
vboxsync |
VMM: MMPhysGCPhys2HCVirt -> PGMPhysGCPhys2HCPtrAssert; deleted MMAllPhys.cpp. |
a7860acddbaa10ea61bbb12772caf00f1ccd8b03 12938 |
|
02-Oct-2008 |
vboxsync |
Comment updates |
0d858e6782c34c1919529b23c80d70729762570d 12937 |
|
02-Oct-2008 |
vboxsync |
SyncPageWorker updates for EPT. |
61fa69e2bc9fc9e7490feed1c020273f3ddb238d 12936 |
|
02-Oct-2008 |
vboxsync |
More EPT updates |
54fe68d807196dc565de24055996e634ee40ee92 12932 |
|
02-Oct-2008 |
vboxsync |
Paging updates in preparation for EPT. |
204418e5b9999c80b3708bef6bbba53d4efebb7c 12894 |
|
02-Oct-2008 |
vboxsync |
Enabled some EPT paging code. |
8ff9db52a08dccde43999d5b9be077469c13650b 12129 |
|
05-Sep-2008 |
vboxsync |
Don't check for hypervisor regions in VT-x/AMD-V mode. |
77bc90b11864bec072f8e6562eda2acd3dd14414 11525 |
|
21-Aug-2008 |
vboxsync |
Added support for PSE-36. |
791aa0a0fbfbf99af3d0c97fec9e7c196d8c183b 11449 |
|
17-Aug-2008 |
vboxsync |
Corrected assertion |
94638b7961a2e43481bed60dcc8c1042dd7c798f 11416 |
|
14-Aug-2008 |
vboxsync |
InvalidatePage: check 4 PDPEs for PAE guest mode. (shouldn't be necessary though) |
7a5a0c6ef590412f909582b752355ce07279245a 11407 |
|
13-Aug-2008 |
vboxsync |
warnings, comments |
e4ea543752422f1139923e3e506c625b0a1827c5 11311 |
|
09-Aug-2008 |
vboxsync |
VMM: ELEMENTS -> RT_ELEMENTS. |
fab9a17099dce8d14fd8d50b547ad4eea377a4dc 10822 |
|
23-Jul-2008 |
vboxsync |
Prepare for EPT. |
bc5ddf7a2288bb337d1eaea5de79513368b116ab 10496 |
|
11-Jul-2008 |
vboxsync |
Attempt 2 |
dfa8248be9f58d614e4c52559c0e6bd29e27f9d5 10495 |
|
11-Jul-2008 |
vboxsync |
AssertCR3 fix for AMD64 paging |
546adac7406114100bd656e199676927e2e38bbb 10343 |
|
07-Jul-2008 |
vboxsync |
Logging update |
41c81f429f890f7e564188fb9551e739bdde819f 10340 |
|
07-Jul-2008 |
vboxsync |
Pool flush handling updates |
65b61798a61dd4c32cce448db1dac70bba8d5cf5 10323 |
|
07-Jul-2008 |
vboxsync |
Clear the shadow page table entries whenever the guest modifies its page tables. This prevents us
from checking all levels in SyncCR3. (amd64 shadow case only) |
69a0f42fa2c531eb2cffb6d6f482d828d9adab34 10321 |
|
07-Jul-2008 |
vboxsync |
Deal with pool flushes directly when possible. |
5dda7f07dab8a954e6c4cf2378b15f921e60d9aa 10320 |
|
07-Jul-2008 |
vboxsync |
Rearranged code to deal with pgm pool tasks during SyncCR3. |
83a2f02f36661f38c8d872a47c6d31040994c70d 10319 |
|
07-Jul-2008 |
vboxsync |
pgmPoolClearAll can be executed in ring 0. |
f48c300ba25e25294885f01b14ee61ac31c69a9a 10318 |
|
07-Jul-2008 |
vboxsync |
There are no virtual handlers in nested paging mode. |
2ddda52bed31c7a1a7904769c603317911f78a3d 10317 |
|
07-Jul-2008 |
vboxsync |
SyncCR3 change for nested paging to deal with pool flushes. |
b7f8ae8dfa02163bc90c1131b990273d18762bd4 10287 |
|
05-Jul-2008 |
vboxsync |
Correction |
70f1b16ef69198fffb29c86835decacede9b4b90 10285 |
|
05-Jul-2008 |
vboxsync |
Corrected protected mode without paging shadow paging. |
9c425bdea5f0991df62922b1584b805a86f2f898 10283 |
|
05-Jul-2008 |
vboxsync |
Backed out 32892-32894 |
85cfe71c2d1640a7d0aa8c8f38ebe198028eef0a 10281 |
|
05-Jul-2008 |
vboxsync |
Disabled user tracking properly. |
f9ca7a3faa0b920674d84abeb170cc9d7c7a62a5 10280 |
|
05-Jul-2008 |
vboxsync |
No need for tracking user pages when guest paging is disabled. |
8c82775e6cb3df46c4ea2f3fba5e024f521cbd81 10270 |
|
05-Jul-2008 |
vboxsync |
Logging update |
daf8a8aaa1a8464106fbf912ad7297ac4c8c233d 10071 |
|
01-Jul-2008 |
vboxsync |
Don't set the no-execute bits in the fake pml4e & pdpte |
e51c95f28de64ec5d434b2339af4fffda43edd73 10068 |
|
01-Jul-2008 |
vboxsync |
Set the accessed bit as well (protected mode without paging; shadow pde) |
9ddf48fac91d47ba87a5255bbd0993297bb16084 10063 |
|
01-Jul-2008 |
vboxsync |
AMD64 updates for nested paging. |
bfe7a8bd955e37013eaddedd99e5adb696d1b470 10013 |
|
30-Jun-2008 |
vboxsync |
AMD64 shadow & real or protected mode without paging combo.
Flush TLB when the EFER msr changes certain bits. |
f9fe97bb14e466908ce846ebc3c6103b7211ef87 10012 |
|
30-Jun-2008 |
vboxsync |
Logging update |
136fa2d5aa26a6272a677616ad4c80f64af07300 10001 |
|
27-Jun-2008 |
vboxsync |
Cleanup |
393b0bc732be29a5637f9bb181c0d4250fdcbaa1 9990 |
|
27-Jun-2008 |
vboxsync |
Some cleanup |
0fef19b085301434b3ff55087ef4f852a32e80bb 9966 |
|
26-Jun-2008 |
vboxsync |
AssertCR3: ignore X86_PTE_PAE_NX as well in big pages (PDE bit is sufficient to override all pte entries) |
d3b142c79c208a407fbbbf0afc568ea5d3b7565b 9942 |
|
26-Jun-2008 |
vboxsync |
PAE fixes |
cdd2418638789661cf6c3628eeed6ac32dc792ff 9941 |
|
26-Jun-2008 |
vboxsync |
Comments |
34266a1c651ebb8d0e6c3af43fb416108d03aba9 9940 |
|
26-Jun-2008 |
vboxsync |
PAE fix |
0e77737b0ba913683e614db11463b31ca67aacbe 9893 |
|
24-Jun-2008 |
vboxsync |
Attempt to fix PAE (can't verify now).
AMD64 paging updates. |
585f64d6f624f9e683321dabeb21b0eb2e6aa473 9890 |
|
24-Jun-2008 |
vboxsync |
amd64 paging updates |
afed5ab737f4aacfae3fe73776f40e989190a7ca 9881 |
|
23-Jun-2008 |
vboxsync |
32 bits iUserTable again (preparing for amd64 changes) |
d02d26dc2fadf0b545fd3bde4fdc743922988bca 9879 |
|
23-Jun-2008 |
vboxsync |
Masks |
6cfc2a96364b95c37975a3688e704507d5e05458 9876 |
|
23-Jun-2008 |
vboxsync |
Logging update |
1b52f11e8f602ad509b4a8653e159641686b5d41 9875 |
|
23-Jun-2008 |
vboxsync |
Typos |
dee51750c0ea39a01a6cd5f4f61705ececc3176f 9871 |
|
23-Jun-2008 |
vboxsync |
Same goes for X86_PTE_G. (big page to pt mapping) |
4a5074a80e22fd7e24d32c53da2aeeb4879af5a9 9870 |
|
23-Jun-2008 |
vboxsync |
Ignore A & D bits in the big page comparison as well. |
21ef63137974c64ee16b2458738ac331dc582d31 9869 |
|
23-Jun-2008 |
vboxsync |
AssertCR3 for long mode. |
b469d1d33d3ce8638db8e238accf810018bba4df 9864 |
|
20-Jun-2008 |
vboxsync |
Logging update |
171814e2bb0e2c94cd1f28b08874a13be7f95088 9863 |
|
20-Jun-2008 |
vboxsync |
Detect PML4E & PDPTE changes in SyncCR3 as well. |
cd5f26c7e9ae8ae39c1fd0f0156ac29fc3de4dfd 9860 |
|
20-Jun-2008 |
vboxsync |
Return immediately in case the PML4E or PDPTE has been changed. |
cc2fd196b9e4cff5fbc0387a5df7921eb7d8159b 9859 |
|
20-Jun-2008 |
vboxsync |
Wrong mask again |
0e40b171b2a96799f797bafa33259366a00c5c6f 9858 |
|
20-Jun-2008 |
vboxsync |
Wrong mask |
2fe20148504a5410c92013227801f655898d83aa 9857 |
|
20-Jun-2008 |
vboxsync |
Updates for amd64 paging |
734dfed68b98b3e0d75f37d497e266399ba155eb 9855 |
|
20-Jun-2008 |
vboxsync |
SyncPage fix for out of sync pde's. |
28eb2212928377140f21a7ddfb122f2ff5e857ef 9778 |
|
17-Jun-2008 |
vboxsync |
InvalidatePage corrections. |
646b2218eea4eb46215f8de030881bc602a996e0 9758 |
|
17-Jun-2008 |
vboxsync |
Wrong assertion |
635e1aeaa8b42d4ea905af5a90bdf25bbe79e492 9757 |
|
17-Jun-2008 |
vboxsync |
Corrections |
62e4411f156db109c8807cb8a64938d3023a22c5 9756 |
|
17-Jun-2008 |
vboxsync |
Wrong masks |
7ec00757c841f7e18ed364dabbc11c2755e9c2e4 9754 |
|
17-Jun-2008 |
vboxsync |
Naming again |
b255ffea54a5368c5f4d383e034f4dd1f774fff7 9753 |
|
17-Jun-2008 |
vboxsync |
Naming |
3ecf9412133496b2aeb090cfd33a286404ec59fb 9752 |
|
17-Jun-2008 |
vboxsync |
PAE & AMD64 paging updates.
- create references to upper level shadow descriptors
- don't flush lower levels if an upper level table entry changes |
21e4801f9fa7e96fe268b310f6203ec98a479c94 9717 |
|
16-Jun-2008 |
vboxsync |
Logging updates |
8c7dd90221b455ab1937039ca542d015370454ea 9711 |
|
16-Jun-2008 |
vboxsync |
Wrong assertion. There's a raw mode with PAE now. |
fc3b8ba312ebc5767b60e245479ab372acc3d678 9692 |
|
13-Jun-2008 |
vboxsync |
Partly backed out previous changeset |
4cad44a5f7cfa7db8d5fb255a1df3b3cd1687853 9690 |
|
13-Jun-2008 |
vboxsync |
Paging updates for amd64. |
04961d8cd0cbe77f5dfc904f3328fe2963171ac4 9684 |
|
13-Jun-2008 |
vboxsync |
Correction |
a85534d1d8f19cbd3c3673485be2a8b9ea68b614 9682 |
|
13-Jun-2008 |
vboxsync |
Long mode paging updates |
46df4404c8dbbf3672e7aae8cd0b2770356e5b73 9620 |
|
11-Jun-2008 |
vboxsync |
Long mode paging updates |
7cbd97f7937248e78edcea80823687c18ffbde1b 9606 |
|
11-Jun-2008 |
vboxsync |
X86_CR4_PSE changes for long mode (bit is ignored there) |
21ffa82e7957c99c78a2c9948cbeed7ca42f6af3 9589 |
|
11-Jun-2008 |
vboxsync |
AMD64 paging updates |
9bc263d82465fc1fa500153d886b691f37b3f034 9570 |
|
10-Jun-2008 |
vboxsync |
AMD64 paging updates |
573ce26cc8c089b7b58aced264d1e9d8c5a5ff3e 9344 |
|
03-Jun-2008 |
vboxsync |
Types and string format specifiers. |
7235946d51c4e7b209b12cefb8a3924660a9d46b 9341 |
|
03-Jun-2008 |
vboxsync |
Format specifiers again... |
78a205e3fc6719d59e8c561b3d287d3a4f879852 9212 |
|
29-May-2008 |
vboxsync |
Major changes for sizeof(RTGCPTR) == uint64_t.
Introduced RCPTRTYPE for pointers valid in raw mode only (RTGCPTR32).
Disabled by default. Enable by adding VBOX_WITH_64_BITS_GUESTS to your LocalConfig.kmk. |
f57ff25c60b8aaa6afdc71b31d8514951c96d862 9041 |
|
21-May-2008 |
vboxsync |
spaces. |
eb5ec011a626159ce61e9aa49175ac8218276ffb 9032 |
|
21-May-2008 |
vboxsync |
Nested paging updates |
f3a64446d7aac8fef8e843d4ca7a087bb41eeada 9021 |
|
21-May-2008 |
vboxsync |
Nested paging updates. Extra paging mode added to prevent illegal changes to the shadow page table. |
134a71c1528b56afe4db843ab63ec5a5b849535b 9001 |
|
21-May-2008 |
vboxsync |
Enabled the PGMMODE_PROTECTED (Guest) & PGMMODE_AMD64 (shadow) combination again. |
6d4f64d7bfb97e9fc7ff3d4bbe0bd16ba5b8a3d5 8739 |
|
09-May-2008 |
vboxsync |
Set the shadow PDE for real and protected mode without paging to read/write & user. |
39a50acba9587ff144e465cacdd41826f50d31c7 8738 |
|
09-May-2008 |
vboxsync |
More logging |
f78748fa5e986ded3741f3ead3cd9534cf1e602b 8659 |
|
07-May-2008 |
vboxsync |
Updates for 64 bits paging.
Removed conditional dirty and accessed bits syncing. Doesn't make sense not to do this. |
9507ebc79a9c49425a43baf5623b10d1de2325d9 8537 |
|
02-May-2008 |
vboxsync |
Spaces |
bc6ee42d72fd0618c818c747f91d016a559645f6 8536 |
|
02-May-2008 |
vboxsync |
Split out the long mode only bits from the PDPE.
Set accessed bits for the PDPE & PML4E |
a9278a54b46f6f0e485c7988be86c79074639568 8535 |
|
02-May-2008 |
vboxsync |
Trying to shut up GCC |
71d0256a894c1b04375c7e3abdf7179df8e11aaa 8533 |
|
02-May-2008 |
vboxsync |
Start of 64 bits paging support |
92f52d885c6c8ceee40b40f1f257a458aafb9d31 8503 |
|
30-Apr-2008 |
vboxsync |
Fall back to the recompiler when the guest tries to access invalid physical memory (e.g. probing RAM or accessing remapping MMIO memory) |
27d45f3d7e23246a7d858dc8670b19d6809a4c9b 8191 |
|
19-Apr-2008 |
vboxsync |
warnings |
1c94c0a63ba68be1a7b2c640e70d7a06464e4fca 8155 |
|
18-Apr-2008 |
vboxsync |
The Big Sun Rebranding Header Change |
0205cdf2a6d8f1d1ee4326a3b6ade00022da161c 8084 |
|
17-Apr-2008 |
vboxsync |
Don't free mappings (PAE mode) |
643c2764713c8c26c6a9ba25d89d0e2489cfb1a9 8079 |
|
17-Apr-2008 |
vboxsync |
PAE bugfix for CR3 mapping |
acceb4bf18ed8d49c5d8034d8ddea1ade50b0d7a 8048 |
|
16-Apr-2008 |
vboxsync |
Indent |
d593c1e5dbcee83ffd495e3f3bcc1eed5995f355 8047 |
|
16-Apr-2008 |
vboxsync |
More logging |
69be17f331f0e51b40688feabc381dda79b0388b 8042 |
|
16-Apr-2008 |
vboxsync |
corrected #if/#else/#endif comments. |
34367d4de514a7f16e9159305acfcf99f36818e5 8034 |
|
16-Apr-2008 |
vboxsync |
Fixed another regression (pae shadow, 32 bits guest) |
fdea543f71872a3ec3909536a4fce37ab7aa3a8b 8033 |
|
16-Apr-2008 |
vboxsync |
Fixed one regression + several PAE related bugs. |
a270916678873bbf2b304d39596d8123b2520fb6 8031 |
|
16-Apr-2008 |
vboxsync |
Accidental debug code |
2296859d5599a60b075166b061e926c811d00e5c 8030 |
|
16-Apr-2008 |
vboxsync |
Another fix for PAE mappings |
a258c107b6d7132f8452a4a93ba85dd60b9d5352 8028 |
|
16-Apr-2008 |
vboxsync |
Precedence |
b102aef4b776d92e786c2cb09b99940703da921c 8027 |
|
16-Apr-2008 |
vboxsync |
Fixed regression in 29680 |
325823d2d3213660ea0581c1be0c667fdf7047e8 8021 |
|
16-Apr-2008 |
vboxsync |
Added missing bits for PAE guest mappings. |
2ed8d64637e35b8117511a02b3d1e60829d08998 8002 |
|
15-Apr-2008 |
vboxsync |
Backed out 29657 |
289f21bb986c6156eab4443a6fc50aa260fa59dd 8001 |
|
15-Apr-2008 |
vboxsync |
PAE monitoring fixes |
a74f394126bf4124e3016240545b10e08c4fba97 7997 |
|
15-Apr-2008 |
vboxsync |
Assertion now only applies to long mode |
f1fc6b69fa0660ad0e600268c1c62e2f9031fe8f 7904 |
|
11-Apr-2008 |
vboxsync |
Additional checks for pgmGstGetPaePDPtr return value |
c313e4a5fce76203c4008c272fdae066261c2281 7885 |
|
10-Apr-2008 |
vboxsync |
Extra check for PAE PDPTR |
b35e836b37dc015e5134379717f8a768152fd7de 7877 |
|
10-Apr-2008 |
vboxsync |
Adjusted assertion |
3381d91e4edaa6d3f1b661fda280beb1d785b53a 7876 |
|
10-Apr-2008 |
vboxsync |
Correction |
9084535d96b2c9d91ab95fdfb0672bffc6735117 7874 |
|
10-Apr-2008 |
vboxsync |
Shutting up the compiler |
f04337b3fe4ee0bdecddf6b7cdb9146df90b725c 7873 |
|
10-Apr-2008 |
vboxsync |
AssertCR3: check the 4 PDPTs too in PAE legacy mode |
3e0c2f619fb8d10f7f9fa689cad76d1b8f902bc7 7859 |
|
10-Apr-2008 |
vboxsync |
Again |
d4594961da0fcba611e559a90f55102af8df195a 7858 |
|
10-Apr-2008 |
vboxsync |
AssertCR3 correction for PAE |
a3dbd14082edea2885045adf99e43977e75dcdee 7845 |
|
09-Apr-2008 |
vboxsync |
Corrected parameters for SyncCR3 & AssertCR3 |
312be3f78f5f0caa406bdc93d3d4e8914ed5eedb 7842 |
|
09-Apr-2008 |
vboxsync |
AssertCR3 for PAE guest mode |
af77549f5851d7262dd1ffae5ec5a05cc302dd54 7822 |
|
09-Apr-2008 |
vboxsync |
Must inform pgmpool about cleared PDs |
e83667fa085bdbfdb5d5d0a72a38364dc498c97d 7820 |
|
09-Apr-2008 |
vboxsync |
Lazy clearing of PAE page directory |
b00e89ebcab8e017bdb514d569e565cd2b1695d6 7819 |
|
09-Apr-2008 |
vboxsync |
Clear PDPT |
1c80e7e49200fed1c00c9e34e6f28639254a5730 7808 |
|
08-Apr-2008 |
vboxsync |
SyncPage: correct NXE bit handling |
bceb174044842b891e0776931f2fea2d312390ff 7807 |
|
08-Apr-2008 |
vboxsync |
Check pointer |
24151b80dba0b79ec2b9a387a966200523148150 7804 |
|
08-Apr-2008 |
vboxsync |
NXE correction |
94681de361a7093acdbc99cee82a42b35889c79e 7802 |
|
08-Apr-2008 |
vboxsync |
NX, r/w, u/s bits in the PDPE are valid in long mode only. |
e8dd72a1287a05f1c2dbcb522939b211acc47fe5 7800 |
|
08-Apr-2008 |
vboxsync |
PAE bugfix |
8d35b5e60c36197e3c52a039bd40f8601b6ce4c0 7797 |
|
08-Apr-2008 |
vboxsync |
Cleanup |
dec1b20d91846834211ebbce8de69128ec1d96d2 7796 |
|
08-Apr-2008 |
vboxsync |
Consistency |
faf6b22ac9299d6a674771df0850f13637754941 7794 |
|
08-Apr-2008 |
vboxsync |
corrected assertion |
382fb7a750711bd5451e16f218fc465dc2ea414a 7793 |
|
08-Apr-2008 |
vboxsync |
More verbose assertion |
81ef6b7c9f45eccc3ce932354c9e9952d6474cf7 7791 |
|
08-Apr-2008 |
vboxsync |
PAE fix for shadow PD index |
3c41ad2ea0d72f91009521afc81047c24268c31f 7747 |
|
04-Apr-2008 |
vboxsync |
PAE correction |
414e42be6c94fc78965792a8279ab56fb99831a7 7743 |
|
04-Apr-2008 |
vboxsync |
And another one |
19a44d9215063c1c7e0cdaec3b0b5ae6e06aa2b1 7742 |
|
04-Apr-2008 |
vboxsync |
Corrected assertion for PAE guest mode |
bda14786cbbda8419eacadb7b96ff8b00e6dd601 7739 |
|
04-Apr-2008 |
vboxsync |
Moved WP check to the right place; fixes stuck boot of XP with hwaccm enable |
9464ac33f05059390b99ceebf88e2e55f97dcda8 7730 |
|
03-Apr-2008 |
vboxsync |
Added CPUMSet/GetGuestEFER.
Corrected NX bit handling. |
22c5bb04202ea3e49a7fb7b02ebce3975f1cb5d5 7728 |
|
03-Apr-2008 |
vboxsync |
Some cleanup.
CheckPageFault: implemented PAE case |
8abe0e0c1cc004c50cf94ff53a7bed1b1df08b45 7719 |
|
03-Apr-2008 |
vboxsync |
pgmGstGetPaePDPtr. |
01ecf57bc89388e4970ae42654894dcbfd64b88a 7718 |
|
03-Apr-2008 |
vboxsync |
Enabled X86_CR0_WP=0 code execution with VT-x & AMD-V |
8f7bc6ad2b7bbcb4b3b96248cd2478e45f2e3b88 7715 |
|
03-Apr-2008 |
vboxsync |
Renamed PDPTR to PDPT.
Added preliminary code for executing code with X86_CR0_WP cleared (disabled). |
1cab4c8b5903a7933e7c31a6b5738e629029c7b3 7702 |
|
02-Apr-2008 |
vboxsync |
Fixed a bunch of #else and #endif comments. The convention here is that on an #else we'll document what the criteria for the following body is, and for #endif the above body. So, the #else comment is the inverse of the #if and the #endif is the same as the #else, or if there isn't any #lese, the #if. |
9ad362a511446bba04cc66bbec1305745b1ca4f3 7698 |
|
02-Apr-2008 |
vboxsync |
Removed redundant check |
8c682a9a078cfd214130816edb813ef5bcac232d 7697 |
|
02-Apr-2008 |
vboxsync |
Bugfix for real & protected mode |
19ec3aeb41eef3f3c5fd02d2412f01811ef6fcf8 7677 |
|
01-Apr-2008 |
vboxsync |
Cleanup |
5476bd9ec30ebad9ed9e39aa8a4ff2c49bede2d7 7676 |
|
01-Apr-2008 |
vboxsync |
Cleaned up.
AMD64 shadow paging is only valid with AMD64 guest paging. Other combinations removed.
Simplified paging #ifdefs. |
8e918e5fbe49284c598cdf54ca25200642e0095d 7668 |
|
31-Mar-2008 |
vboxsync |
Cleaned up |
92b424059f84323c53f6f31a90c77cdd7f5760f3 7667 |
|
31-Mar-2008 |
vboxsync |
More comments + PAE updates |
11c0410636b200260d9e040d2f0de09592c55a66 7666 |
|
31-Mar-2008 |
vboxsync |
Put back the original PAE defines. AMD differs from Intel wrt valid pae & long mode physical address bits. |
90a2b54e0f5fb9db3313015676a65d877cac1c09 7664 |
|
31-Mar-2008 |
vboxsync |
Comment update |
30ea0ddca87750d8fd5719861a91884a8de246d9 7662 |
|
31-Mar-2008 |
vboxsync |
Clarified code |
df9dc33be49acc653ac9aa62e64c9c56e6f05a27 7657 |
|
31-Mar-2008 |
vboxsync |
Renamed 4M to BIG |
9b491eb228b51d7ab6de858ae1f4e011b3d9527d 7656 |
|
31-Mar-2008 |
vboxsync |
Use GST_PDE4M_PG_MASK |
b257884bf5d3c5b54ea68b86d58e4060d8f0ddd8 7655 |
|
31-Mar-2008 |
vboxsync |
PAE fixes & changes |
7784f1038e2518d2f816b9d7a43068d5dbdbf366 7642 |
|
31-Mar-2008 |
vboxsync |
Additional cleanup |
f3551b3ca8c37fca817365c30827c8c6ed8fcd3e 7629 |
|
28-Mar-2008 |
vboxsync |
Initial cleanup for PAE |
bd1ef32f3269ea52098ba14a0aa193a74f932447 7577 |
|
26-Mar-2008 |
vboxsync |
~0U -> NIL_RTGCPHYS |
c134d031907a71e73152acdf08daf4c3ca28fe84 6927 |
|
12-Feb-2008 |
vboxsync |
Converted MM_RAM_FLAGS_VIRTUAL_HANDLER, MM_RAM_FLAGS_VIRTUAL_WRITE
and MM_RAM_FLAGS_VIRTUAL_ALL into a two bit state variable in PGMPAGE.
I've checked this trice because, like last time, bugs may have odd
sideeffects and hide for a while before showing up. Hope I got this
right (unlike for phys).
Fixed a regression from the MM_RAM_FLAGS_PHYSICAL in the physical read/write code. |
4f7e25c6efee4ff5419050d168f6de8333609c0a 6913 |
|
12-Feb-2008 |
vboxsync |
Separated the PGMVIRTHANDLERTYPE_HYPERVISOR kind of handlers from the others since to simplify the code a bit. |
88d5f0bc071bee7aabc1019c9bec949780abe7d2 6911 |
|
11-Feb-2008 |
vboxsync |
Removed the EIP (PGMVIRTHANDLERTYPE_EIP) kind of virtual access handlers. This type have never been used, isn't working right, and is just messing up the code. |
ffe7461c63ca36e2d358f4882963e23c1dc504fa 6902 |
|
11-Feb-2008 |
vboxsync |
Converted MM_RAM_FLAGS_PHYSICAL_HANDLER, _WRITE, _ALL and _TEMP_OFF into
a 2-bit state field (u2HandlerPhysStateX). I've tripple checked this change,
but if I overlooked something real odd stuff might happen... |
c20b837b2c912dd4dcaaa676e15acd2b230108f0 6829 |
|
06-Feb-2008 |
vboxsync |
Addressed the R0/R3 issues with the PGMRAMRANGE structure. |
a16eb14ad7a4b5ef91ddc22d3e8e92d930f736fc 5999 |
|
07-Dec-2007 |
vboxsync |
The Giant CDDL Dual-License Header Change. |
5d5a16e354437740c528da44783f21bf1a3a194a 5663 |
|
10-Nov-2007 |
vboxsync |
Inverted test |
63a13eb21fc0194368456e02048dd6baebb3ca24 5662 |
|
10-Nov-2007 |
vboxsync |
Initial changes for guest PAE support |
a5e542632108d84ba1777c44a9c0bcbb689288c9 5629 |
|
05-Nov-2007 |
vboxsync |
remove deprecated macros |
3e6c6998d1dfeded8b9a23f5aa94ad63e9a681d9 5605 |
|
01-Nov-2007 |
vboxsync |
BIT => RT_BIT, BIT64 => RT_BIT_64. BIT() is defined in Linux 2.6.24 |
5a3ac2e10bd820f740b06f6ea59cd5063d6e4a8f 5006 |
|
24-Sep-2007 |
vboxsync |
|
c04d56e828cd419634eaad25265678196c1c72cd 4999 |
|
24-Sep-2007 |
vboxsync |
More logging |
340d7413d3aba0ad850ce5f164045e7df9aedcff 4992 |
|
24-Sep-2007 |
vboxsync |
LogFlow instead |
8256e2948cf3dde3ae37aaaffce8dbe682e5705e 4991 |
|
24-Sep-2007 |
vboxsync |
More logging |
223cf005b18af2c21352a70693ebaf0582f68ebc 4977 |
|
22-Sep-2007 |
vboxsync |
Backed out most of 24659. |
19b725c530eb49600728765de7ed451cbe290740 4917 |
|
20-Sep-2007 |
vboxsync |
Don't trigger breakpoints in ring 0. (AssertRelease & co) |
d92790f0563ea59b9c6f8538b8866a4259d28e06 4788 |
|
14-Sep-2007 |
vboxsync |
Logging level change |
56c9a42f56ef04b3f01d84e7ea8277f255899e72 4665 |
|
10-Sep-2007 |
vboxsync |
Moved some of the odd address conversion routines to PGMR3Dbg just to get them out of the way. |
aaeb2e2f6ed5b164f1dec9a16a7adeb84f64cf31 4620 |
|
08-Sep-2007 |
vboxsync |
The initial PGMRAMRANGE::aHCPhys -> PGMRAMRANGE::aPages (PGMPAGE) conversion. |
c98fb3e16fcd571a790eab772c0c66173d225205 4071 |
|
07-Aug-2007 |
vboxsync |
Biggest check-in ever. New source code headers for all (C) innotek files. |
6b626ce1907577d6505296ebcbc030c9b33691aa 3185 |
|
20-Jun-2007 |
vboxsync |
some minor comment cleanup. |
c97989161fbe75bc14cea477a5443bbf474dd3ad 2981 |
|
01-Jun-2007 |
vboxsync |
InnoTek -> innotek: all the headers and comments. |
375cd37eacaaf8bbcc0d18c0f6443856bcd0c635 2701 |
|
17-May-2007 |
vboxsync |
Stupid typo |
622f96e6144fecfd77ebef1f57a556ea5cfeab8a 2559 |
|
09-May-2007 |
vboxsync |
Allow ring-3 only access handlers (phys). |
247c3f8cda8ea62aed0cff94a1987aa85d9a42df 2276 |
|
20-Apr-2007 |
vboxsync |
VerifyAccessSyncPage: shadow PD might not be present. |
ac27faaad009fc46b7be690e6cefe9a24deba27a 2274 |
|
20-Apr-2007 |
vboxsync |
Pass on rc. |
88e43a2222e2db89f84971ae2682253689c09055 2273 |
|
20-Apr-2007 |
vboxsync |
Real and protected mode (no paging) implemenation of VerifyAccessSyncPage. |
d1011b3f539ad819b5e2c1c3d8152e7113725bf6 2270 |
|
20-Apr-2007 |
vboxsync |
Stricter pointer typechecking. (R0 vs R3) |
713a876c17d2befe27e3c2910491dfb4a5e661eb 2203 |
|
19-Apr-2007 |
vboxsync |
prevent warning |
b95f77de8893e184e5525cf072b74721ed06485c 2149 |
|
18-Apr-2007 |
vboxsync |
Set user bit in shadow page & pd when in non-paging mode. |
59bd79968969d18bda6bfb098ae9ccc697287996 2137 |
|
17-Apr-2007 |
vboxsync |
Drop page flags consistency check in non-paging mode. |
2a8d3fa1b57112afd8f32694cb07da05187ca341 2120 |
|
17-Apr-2007 |
vboxsync |
Real and protected mode without paging changes. Currently not yet used. |
34deeaa787527c3e7de4484db5bf9f838ae29452 2078 |
|
13-Apr-2007 |
vboxsync |
Removed obsolete code |
f37011f6d4251c4af0c6ee694cc170d072cea0a4 1908 |
|
04-Apr-2007 |
vboxsync |
Backed out part of 20091 to check if that's the cause of the major performance regression seen on testboxlin. |
1d29f0f00aa5c78b32e89bca759ab70390f75c32 1881 |
|
03-Apr-2007 |
vboxsync |
Unused warning. |
0d135b99b04c819ef20e6bde8b2bb10435d50150 1866 |
|
02-Apr-2007 |
vboxsync |
Disabled previous csam changes. |
246d1cc1f91258f62314afeedcb50f68b6894c6c 1865 |
|
02-Apr-2007 |
vboxsync |
InvalidatePage: always make page non-present (possibly dangerous)
Page fault handler: updates for extra code scanning. |
508878f8bf929589c2908f77999a8b1ed0a63dc1 1839 |
|
30-Mar-2007 |
vboxsync |
style |
baaa8f2deb495b8136c512d3efe6321817cacd1b 1828 |
|
30-Mar-2007 |
vboxsync |
Cleaned up cpl checking. |
7354012663128de04d0068aea698106a10204edc 1823 |
|
30-Mar-2007 |
vboxsync |
Extra safety check. |
8f6d1a96a3c1d44545b0ca3d791a0e10a2947280 1821 |
|
30-Mar-2007 |
vboxsync |
Extra check for hypervisor region writes. |
9810d5a703e2de217dc51c6169c96f8e9d70cd44 1620 |
|
22-Mar-2007 |
vboxsync |
All rings are allowed with hardware acceleration |
092b6023a33fff84301ea050d1249cfcf54b224a 1359 |
|
09-Mar-2007 |
vboxsync |
SELM function changes for v86 mode code.
CPL check fixes for V86 mode code. |
3cf9030cb3aba5927c755c3e57e8ef4e2bea4ce5 1158 |
|
02-Mar-2007 |
vboxsync |
Corrected assertion |
ad3c1f0d63e8e53da0b9d747ce4d2f6557e65380 1090 |
|
28-Feb-2007 |
vboxsync |
Corrected assertions for v86 mode |
ccbc8c0ae6f303b4fdf1484d838fd23b17e67334 851 |
|
12-Feb-2007 |
vboxsync |
Refresh HCPhys. |
d8c5aaee199e6169571540141186173301ad7d1f 850 |
|
12-Feb-2007 |
vboxsync |
Missing case where we should check that the guest RAM was indeed allocated. |
308abff56ead3a7df68c25111148318844021b47 838 |
|
12-Feb-2007 |
vboxsync |
GC Phys to HC virt conversion changes for dynamic RAM allocation. |
fb5e37303b228a79c05cabfce2fe0fedfe32ed8a 23 |
|
15-Jan-2007 |
vboxsync |
string.h & stdio.h + header cleanups. |
677833bc953b6cb418c701facbdcf4aa18d6c44e 1 |
|
01-Jan-1970 |
vboxsync |
import |