This PCD holds the address mask for page table entries when memory
encryption is enabled on AMD processors supporting the Secure Encrypted
Virtualization (SEV) feature.
The mask is applied when page tables entries are created or modified.
CC: Jeff Fan <jeff.fan@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Brijesh Singh <brijesh.singh@amd.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Leo Duran <leo.duran@amd.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Platform PEI may add LOCAL APIC memory mapped space into
EFI_HOB_MEMORY_ALLOCATION. Or platform may allocate this range before.
So, we skip AllocateMemorySpace()'s return status checking. Instead, we add one
DEBUG message for possible trace.
https://bugzilla.tianocore.org/show_bug.cgi?id=390
This updating is suggested by Ersek's comments at
https://www.mail-archive.com/edk2-devel@lists.01.org/msg22585.html
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Star Zeng <star.zeng@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Current CpuGetTimerValue() implementation return hard code TimerPeriod value. We
could calculate the actual TimerPeriod value over period of time (100us) at the
first time invoking CpuGetTimerValue() and save the TimerPeriod value into one
global variable to avoid delay at the next CpuGetTimerValue() invoking.
https://bugzilla.tianocore.org/show_bug.cgi?id=382
Cc: Feng Tian <feng.tian@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Feng Tian <feng.tian@Intel.com>
Local APIC memory mapped space should be added into GCD and be allocated.
Otherwise, UEFI firmware cannot get correct memory map for it. For example,
SMM profile feature needs to get the completed MMIO map to protect them.
v2:
Consume AddMemoryMappedIoSpace() to handle the case that Local APIC
memory space has already been added before.
https://bugzilla.tianocore.org/show_bug.cgi?id=390
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
Add memory attribute setting in CpuArch protocol.
Previous SetMemoryAttributes() API only supports cache attribute setting.
This patch updated SetMemoryAttributes() API to support memory attribute
setting by updating CPU page table.
Cc: Jeff Fan <jeff.fan@intel.com>
Cc: Michael Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Cc: Jeff Fan <jeff.fan@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
If CPU Bist data is not zero, we will report Status code. But there is one bug
that will report each processor's status code duplicated with NumberOfData
times. This fix is to exchange the loop order on NumberOfData and
mNumberOfProcessors. It could make sure the report status code only once for
each processor.
Cc: Michael Kinney <michael.d.kinney@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Giri P Mudusuru <giri.p.mudusuru@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Feng Tian <feng.tian@Intel.com>
v5:
If PcdCpuMaxLogicalProcessorNumber is set to 1 on UP system,
MpInitLibInitialize() will be invoked. This is one bug, we need to
call MpInitLibInitialize() always and get the BSP information.
Just to remove PcdCpuMaxLogicalProcessorNumber() consuming from this
driver.
Cc: Michael Kinney <michael.d.kinney@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Giri P Mudusuru <giri.p.mudusuru@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Michael Kinney <michael.d.kinney@intel.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Michael Kinney <michael.d.kinney@intel.com>
Consume MP Initialize library to produce CPU MP Protocol services to simply the
code.
v4:
1. Update CpuDxe.c file header to mention it produces CPU Arch protocol.
2. Update BistData type from UINT32 to EFI_HEALTH_FLAG.
3. Move some header location from CpuMp.h to CpuDxe.h.
v3:
1. Move the code Consume MpInitLib APIs to produce CPU MP Protocol from patch
#40 to this patch.
2. Add DxeMpInitLib.inf in DSC file
Cc: Michael Kinney <michael.d.kinney@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Giri P Mudusuru <giri.p.mudusuru@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Michael Kinney <michael.d.kinney@intel.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Michael Kinney <michael.d.kinney@intel.com>
SetMemoryAttributes() will sync BSP's MTRRs settings to all APs by StartupAllAPs
service in serial mode. It may caused much performance impact if there are too
much processors in system. This update is to invoke StartupAllAps in parallel
mode. IA32 SDM does suggest to program MTRRs in parallel mode.
Cc: Michael Kinney <michael.d.kinney@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
The BaseTools/Scripts/ConvertMasmToNasm.py script was used to convert
Ia32/CpuAsm.asm to Ia32/CpuAsm.nasm
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Liming Gao <liming.gao@intel.com>
To convert these files I ran:
$ python3 BaseTools/Scripts/ConvertUni.py UefiCpuPkg
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19264 6f19259b-4bc3-4df7-8a09-765794883524
The CpuDxe driver may run before the gEfiMetronomeArchProtocolGuid
protocol is installed. gBS->Stall does not work until this arch
protocol is installed.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18914 6f19259b-4bc3-4df7-8a09-765794883524
The PiSmmCpuDxeSmm module makes some assumptions about GDT selectors
that are based on the GDT layout from the DxeIplPeim. For example,
the protected mode entry code and (where appropriate) the long mode
entry code in the UefiCpuPkg/PiSmmCpuDxeSmm/*/MpFuncs.* assembly
files, which are used during S3 resume, open-code segment selector
values that depend on DxeIplPeim's GDT layout.
This updates the CpuDxe module to use the same GDT layout as the
DxeIplPeim. This enables modules that are dispatched after
CpuDxe to find, and potentially save and restore, a GDT layout that
matches that of DxeIplPeim. The DxeIplPeim has a 2 GDT entries for
data selectors that are identical. These are LINEAR_SEL (GDT Offset
0x08)and LINEAR_DATA64_SEL (GDT offset 0x30). LINEAL_SEL is used for
for IA32 DXE and the LINEAR_DATA64_SEL is used for X64 DXE. This
duplicate data selector was added to the CpuDxe module to keep the
GDT and all selectors consistent.
Using a consistent GDT also improves debug experience.
Reported-by: Laszlo Ersek <lersek@redhat.com>
Analyzed-by: Laszlo Ersek <lersek@redhat.com>
Link: http://article.gmane.org/gmane.comp.bios.edk2.devel/3568
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18710 6f19259b-4bc3-4df7-8a09-765794883524
The Quark_EDKII_v1.1.0/IA32FamilyCpuBasePkg/CpuArchDxe
driver applies any MTRR changes to APs, if the
EFI_MP_SERVICES_PROTOCOL is available. We should do the same.
Additionally, the broadcast should occur at MP startup as well,
not only when MTRR settings are changed. The inspiration is
taken from
Quark_EDKII_v1.1.0/IA32FamilyCpuBasePkg/CpuMpDxe/
(see the EarlyMpInit() function and its call sites in
"ProcessorConfig.c").
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Michael Kinney <michael.d.kinney@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18632 6f19259b-4bc3-4df7-8a09-765794883524
Address a race condition in first call to StartupAllAPs() with
SingleThread set to TRUE in the MP initialization. If the APs
have not entered their idle loop before StartupAllAPs() is called,
then some of the APs will be in an unexpected state, and
StartupAllAPs() will hang. This is the hang condition that is
only seen when SingleThread parameter is set to TRUE and
StartupAllAPs() is called very shortly after mAPsAlreadyInitFinished
is set to TRUE that releases the APs to complete their initialization.
An internal function has been added to check if all APs are in the
sleeping state in their idle loop. On the first call to
StartupAllAPs(), this internal function is used in a loop to make
sure the APs are in a state that is compatible with the use of
StartupAllAPs(). PcdCpuApInitTimeOutInMicroSeconds is used as the
maximum wait time for the APs to enter their idle loop. If all
APs have not entered their idle loop within the timeout, then
an ASSERT() is generated.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Cc: Jeff Fan <jeff.fan@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18631 6f19259b-4bc3-4df7-8a09-765794883524
Use PcdCpuApInitTimeOutInMicroSeconds instead of hardcoded 100ms for
the time to wait for all APs to respond to first INIT SIPI SIPI
wake request.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Cc: Jeff Fan <jeff.fan@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18630 6f19259b-4bc3-4df7-8a09-765794883524
Only perform AP detection if PcdCpuMaxLogicalProcessorNumber > 1
Only free AP related structures of they were allocated
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18629 6f19259b-4bc3-4df7-8a09-765794883524
when stack NX has been enabled for BSP.
DxeIpl may have enabled Execute Disable for BSP,
APs need to get the status and sync up the settings,
otherwise EFI_MP_SERVICES_PROTOCOL->StartupAllAPs
may not work.
Got positive comments and test result from Laszlo
for the early draft patch, thanks.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Star Zeng <star.zeng@intel.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18191 6f19259b-4bc3-4df7-8a09-765794883524
Should check NumberOfData value instead of treat its value as one BOOLEAN type.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Shuming Qiu <shuming.qiu@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@17672 6f19259b-4bc3-4df7-8a09-765794883524
Get CPU BIST information from gEfiSecPlatformInformation2PpiGuid or
gEfiSecPlatformInformationPpiGuid Guided HOB and update the CPU healthy status
for CPU MP Service.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@17641 6f19259b-4bc3-4df7-8a09-765794883524
The DDK3790 tool chain fails when the PAUSE instruction is assembled:
error A2085: instruction or register not accepted in current CPU mode The solution is to use the .686 directive along with the .xmm directive.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Scott Duplichan <scott@notabs.org>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@17134 6f19259b-4bc3-4df7-8a09-765794883524
For avoid the compiler optimizing the code, we let Parameter and Procedure in CpuData volatile.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@17024 6f19259b-4bc3-4df7-8a09-765794883524
Add a new sleeping state for APs, when no procedure execution, put AP to sleep. when need to execute
procedure, only need to wake up this AP by sent SIPI.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@17023 6f19259b-4bc3-4df7-8a09-765794883524
when gBS->ExitBootServices() is called, the APs should avoid to access
the unsafed buff datas which were allocated by boot services.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
Tested-by: Gabriel Somlo <somlo@cmu.edu>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@16397 6f19259b-4bc3-4df7-8a09-765794883524
This sequence should happen:
* CpuMp.c: Allocate a stack for the APs
* ApStartup.c: Send Start IPI to wake APs in 16-bit real mode
* MpAsm.S: AP enters CpuDxe driver code without stack
- AP grabs a lock
- AP sets up stack
- AP calls CpuMp.c:ApEntryPointInC
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@16370 6f19259b-4bc3-4df7-8a09-765794883524
If timeout expires before AP returns from Procedure, the AP should
be terminated, we introduce ResetApStackLess() to send init IPI
to let AP exit Procedurce and re-available.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@16366 6f19259b-4bc3-4df7-8a09-765794883524
Due to the implementation of AcquireSpinLock() is not MP safe,
so we should use AcquireSpinLockOrFail directly instead.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Chen Fan <chen.fan.fnst@cn.fujitsu.com>
Reviewed-by: Jeff Fan <jeff.fan@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@16357 6f19259b-4bc3-4df7-8a09-765794883524
Leo Duran
Jeff Fan
Jiewen Yao
Dandan Bi
Liming Gao
Jordan Justen
Michael Kinney
Laszlo Ersek
Star Zeng
Scott Duplichan
Chen Fan