Nagaraj Hegde [Fri, 6 May 2016 10:20:00 +0000 (18:20 +0800)]
NetworkPkg:HttpDxe: Code changes to support HTTP PUT/POST operations
Code changes enables HttpDxe to handle PUT/POST operations.
EfiHttpRequest assumes "Request" and "HttpMsg->Headers" can
never be NULL. Also, HttpResponseWorker assumes HTTP Reponse
will contain headers. We could have response which could contain
only a string (HTTP 100 Continue) and no headers. Code changes
tries to do-away from these assumptions, which would enable
HttpDxe to support PUT/POST operations.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Hegde, Nagaraj P nagaraj-p.hegde@hpe.com Reviewed-By: Wu Jiaxin <jiaxin.wu@intel.com> Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
Yonghong Zhu [Thu, 12 May 2016 06:22:56 +0000 (14:22 +0800)]
BaseTools/GenFw: enhance to use Magic Field to identify the image
Original use the File Header Machine Field to identify
EFI_IMAGE_OPTIONAL_HEADER32 or EFI_IMAGE_OPTIONAL_HEADER64, it cannot
correctly handle EBC arch PE32 image.
Yonghong Zhu [Wed, 11 May 2016 08:09:26 +0000 (16:09 +0800)]
BaseTools/GenFds: enhance INF built arch filter
The bug is use FILE_GUID override to build the same module more than
once, GenFds report warning "xxx NOT found in DSC file; Is it really
a binary module?". The root cause is the module path with FILE_GUID
overridden has the file name FILE_GUIDmodule.inf, then
PlatformDataBase.Modules use FILE_GUIDmodule.inf as key which cause
__GetPlatformArchList__ return empty.
Yonghong Zhu [Wed, 11 May 2016 05:46:44 +0000 (13:46 +0800)]
BaseTools: Fix bug to not mix comment into Asbuilt inf Depex section
in the generated Asbuilt inf would include the driver's complete
dependency expression, and it would be wrote as comment format. Original
bug is mix the depex expression with real comment in the depex section.
this patch is ignore the real comment, and list the depex expression.
Star Zeng [Fri, 13 May 2016 05:01:28 +0000 (13:01 +0800)]
DuetPkg FSVariable: return error for empty str VariableName to GetVariable
Current GetVariable implementation will return the first variable for
empty str VariableName, it is because GetVariable and GetNextVariablename
are sharing same function FindVariable.
But UEFI sepc defines SetVariable that If VariableName is an empty string,
then EFI_INVALID_PARAMETER is returned, that means an empty string variable
could never be set successfully, so GetVariable should return error for
empty string VariableName.
Cc: Jiewen Yao <jiewen.yao@intel.com> Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Star Zeng <star.zeng@intel.com> Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Star Zeng [Fri, 13 May 2016 05:00:01 +0000 (13:00 +0800)]
MdeModulePkg Variable: return error for empty str VariableName to GetVariable
Current GetVariable implementation will return the first variable for
empty str VariableName, it is because GetVariable and GetNextVariablename
are sharing same function FindVariable.
But UEFI sepc defines SetVariable that If VariableName is an empty string,
then EFI_INVALID_PARAMETER is returned, that means an empty string variable
could never be set successfully, so GetVariable should return error for
empty string VariableName.
Cc: Jiewen Yao <jiewen.yao@intel.com> Cc: Feng Tian <feng.tian@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Star Zeng <star.zeng@intel.com> Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Maurice Ma [Fri, 13 May 2016 18:43:23 +0000 (11:43 -0700)]
CorebootPayloadPkg: Use generic SerialDxe driver
Use generic SerialDxe driver in MdeModulePkg instead of the one
in CorebootModulePkg. By doing this the reference for
PciSioSerialDxe driver will also be removed from DSC and FDF file.
Cc: Prince Agyeman <prince.agyeman@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Maurice Ma <maurice.ma@intel.com> Reviewed-by: Lee Leahy <leroy.p.leahy@intel.com>
Jiewen Yao [Fri, 13 May 2016 05:00:53 +0000 (13:00 +0800)]
Add IntelFsp2Pkg and IntelFsp2WrapperPkg.
Add FSP2.0 support.
This series of patch is to support FSP2.0 specification at
https://firmware.intel.com/sites/default/files/FSP_EAS_v2.0_Draft%20External.pdf
Some major updates include:
1) One FSP binary is separated to multiple components:
FSP-T, FSP-M, FSP-S, and optional FSP-O.
Each component has its own configuration data region.
2) All FSP-APIs use same UPD format - FSP_UPD_HEADER.
3) Add EnumInitPhaseEndOfFirmware notifyphase.
4) FSP1.1/FSP1.0 compatibility is NOT maintained.
5) We also add rename Fsp* to FspWrapper* in IntelFsp2WrapperPkg,
to indicate that it is for FspWrapper only.
IntelFspPkg and IntelFspWrapperPkg will be deprecated.
The new Intel platform will follow FSP2.0 and use IntelFsp2Pkg
and IntelFsp2WrapperPkg.
The old platform can still use IntelFspPkg and IntelFspWrapperPkg
for compatibility consideration.
Cc: Giri P Mudusuru <giri.p.mudusuru@intel.com> Cc: Maurice Ma <maurice.ma@intel.com> Cc: Ravi P Rangarajan <ravi.p.rangarajan@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Jiewen Yao <jiewen.yao@intel.com> Reviewed-by: Giri P Mudusuru <giri.p.mudusuru@intel.com> Reviewed-by: Maurice Ma <maurice.ma@intel.com> Reviewed-by: Ravi P Rangarajan <ravi.p.rangarajan@intel.com>
Eric Dong [Thu, 12 May 2016 05:25:26 +0000 (13:25 +0800)]
MdeModulePkg UiApp: change code for easy customization.
Change file arrangement to let user easy customize the front page UI.
After this change:
1. UiCustomizeFrontPage function in FrontPageCustomizedUi.c use to
let user customize front page menus. UiFrontPageCallbackHandler
function use to handle the callback for the added menus.
2. FrontPageCustomizedUiSupport.c file used to add support functions
used by FrontPageCustomizedUi.c file.
3. After this change, only above two files need to be updated if front page
menus needed to be changed.
Dong, Eric [Thu, 12 May 2016 08:32:17 +0000 (16:32 +0800)]
BootMaintenanceManagerUiLib: change code for customization.
Change file arrangement to let user easy customize the BMM UI.
After this change:
1. UiCustomizeBMMPage function in BootMaintenanceManagerCustomizedUi.c
use to let user customize BMM first page menus. UiBMMCallbackHandler
function use to handle the callback for the added menus.
2. BootMaintenanceManagerCustomizedUiSupport.c file used to add support
functions used by BootMaintenanceManagerCustomizedUi.c file.
3. After this change, only above two files need to be updated if BMM
first page menus needed to be changed.
Nagaraj Hegde [Thu, 5 May 2016 06:59:03 +0000 (14:59 +0800)]
MdeModulePkg:DxeHttpLib: Add checks in HttpGenRequestMessage API
HttpGenRequestMessage assumes that HTTP message would always
contain a request-line, headers and an optional message body.
However, subsequent to a HTTP PUT/POST request, HTTP requests
would contain just the message body. This patch supports
creation of such request messages with additional checks.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Hegde, Nagaraj P <nagaraj-p.hegde@hpe.com> Reviewed-By: Wu Jiaxin <jiaxin.wu@intel.com> Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
Leahy, Leroy P [Thu, 12 May 2016 23:05:15 +0000 (16:05 -0700)]
CorebootPayloadPkg: Add OHCI driver
Add the USB OHCI driver from revision 24ca2f35 of QuarkSocPkg.
Change-Id: Ie7aa0bc47d4ff06adc57976a5efb0e40ce4e1673
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Lee Leahy <leroy.p.leahy@intel.com> Reviewed-by: Maurice Ma <maurice.ma@intel.com>
Leahy, Leroy P [Tue, 10 May 2016 22:34:48 +0000 (15:34 -0700)]
CorebootPayloadPkg: Add SD/eMMC support
Add SD and eMMC DXE driver support to CorebootPayloadPkg.
Change-Id: Ibfd3a2cc32a653ce51e38d9157ea3c6da25a5474
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Lee Leahy <leroy.p.leahy@intel.com> Reviewed-by: Maurice Ma <maurice.ma@intel.com> Reviewed-by: Prince Agyeman <prince.agyeman@intel.com>
Leahy, Leroy P [Tue, 10 May 2016 22:34:36 +0000 (15:34 -0700)]
CorebootPayloadPkg: Set the proper Shell file GUID
Set the proper Shell file GUID so that the BDS transfers control to the
Shell.
Change-Id: I816636a340bbe2f76ac1973b9cb685084c4f88a0
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Lee Leahy <leroy.p.leahy@intel.com> Reviewed-by: Maurice Ma <maurice.ma@intel.com> Reviewed-by: Prince Agyeman <prince.agyeman@intel.com>
Leahy, Leroy P [Tue, 10 May 2016 22:34:22 +0000 (15:34 -0700)]
CorebootPayloadPkg: Use correct BaseSerialPortLib16550
Use the BaseSerialPortLib16550 which sets RTS and DTR during
initialization. This fixes the mis-matched flow control issue when
the flow control signals are connected between the host and target
and the host has flow control enabled.
Change-Id: I3505e129b2de3c5c17fff23c62553f15cd892dca
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Lee Leahy <leroy.p.leahy@intel.com> Reviewed-by: Maurice Ma <maurice.ma@intel.com> Reviewed-by: Prince Agyeman <prince.agyeman@intel.com>
Leahy, Leroy P [Tue, 10 May 2016 22:34:11 +0000 (15:34 -0700)]
CorebootPayloadPkg: Assume no PCI serial devices
Set the vendor to 0xffff which indicates the end of the list.
Change-Id: If6475e04d3675f0a932571a85d1dd3f301416b6a
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Lee Leahy <leroy.p.leahy@intel.com>
eviewed-by: Maurice Ma <maurice.ma@intel.com> Reviewed-by: Prince Agyeman <prince.agyeman@intel.com>
Leahy, Leroy P [Tue, 10 May 2016 22:33:57 +0000 (15:33 -0700)]
CorebootPayloadPkg: Use DOS line endings
Convert to using DOS line endings.
Change-Id: Ie2f148867d9b2b386d556583afb6716ec21399e9
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Lee Leahy <leroy.p.leahy@intel.com> Reviewed-by: Maurice Ma <maurice.ma@intel.com>
Ard Biesheuvel [Wed, 11 May 2016 08:38:47 +0000 (10:38 +0200)]
ArmPkg/ArmLib: don't invalidate entire I-cache on range operation
Instead of cleaning the data cache to the PoU by virtual address and
subsequently invalidating the entire I-cache, invalidate only the
range that we just cleaned. This way, we don't invalidate other
cachelines unnecessarily.
Gary Lin [Wed, 11 May 2016 08:40:49 +0000 (16:40 +0800)]
OvmfPkg/PlatformBootManagerLib: Postpone the shell registration
We currently register the shell before creating the boot options for
the block devices and the network devices, so the boot manager boots
into the internal shell if the user doesn't specify the boot order.
However, Xen doesn't support fw_cfg, so there is no way to change the
boot order with the external command, and the firmware will always
boot into the internal shell if the user doesn't interfere the boot
process.
This patch postpones the shell registration after MdeModulePkg/BDS
creates all the boot options for the block and network devices, so
that firmware will try to boot the block/network devices first.
Cc: Laszlo Ersek <lersek@redhat.com> Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Gary Lin <glin@suse.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Michael Brown [Wed, 11 May 2016 21:50:09 +0000 (22:50 +0100)]
EmbeddedPkg/Lan9118Dxe: Do not return uninitialised TxBuff
Conform to the specification for GetStatus(), which states that "if
there are no transmit buffers to recycle and TxBuf is not NULL, *TxBuf
will be set to NULL".
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Michael Brown <mcb30@ipxe.org> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Laszlo Ersek [Mon, 9 May 2016 11:32:27 +0000 (19:32 +0800)]
OvmfPkg/QemuBootOrderLib: adapt Q35 SATA PMPN to UEFI spec Mantis 1353
On the Q35 machine type of QEMU, there is no port multiplier connected to
the on-board SATA controller. Therefore the AtaAtapiPassThru driver update
for Mantis ticket 1353 <https://mantis.uefi.org/mantis/view.php?id=1353>
changes the middle number (the Port Multiplier Port Number) in the Sata()
device path nodes from 0x0 to 0xFFFF.
Adapt the translation from OpenFirmware in QemuBootOrderLib.
(Note: QemuBootOrderLib is deprecated at this point (see USE_OLD_BDS in
the DSC files), but until we remove it, it should be kept in sync with
QemuNewBootOrderLib.)
Laszlo Ersek [Mon, 9 May 2016 11:32:26 +0000 (19:32 +0800)]
OvmfPkg/QemuNewBootOrderLib: adapt Q35 SATA PMPN to UEFI spec Mantis 1353
On the Q35 machine type of QEMU, there is no port multiplier connected to
the on-board SATA controller. Therefore the AtaAtapiPassThru driver update
for Mantis ticket 1353 <https://mantis.uefi.org/mantis/view.php?id=1353>
changes the middle number (the Port Multiplier Port Number) in the Sata()
device path nodes from 0x0 to 0xFFFF.
Adapt the translation from OpenFirmware in QemuNewBootOrderLib.
MdeModulePkg Ata: Use the new (incompatible) PortMultiplierPort semantics
The Mantis ticket 1353 <https://mantis.uefi.org/mantis/view.php?id=1353>
and Mantis ticket 1472 <https://mantis.uefi.org/mantis/view.php?id=1472>
updated the description of the port multiplier port number parameter in
SATA Device Path Node and ATA Pass-Through Protocol.
Now, this parameter should be set to 0xFFFF instead of 0 to indicate that
an ATA device is directly attached on the controller port.
Please note that this is an incompatible change. The consumer of SATA
device path or ATA_PASS_THRU needs to re-examine its usage to follow UEFI
2.5 mantis 1353 and 1472.
MdePkg Protocol/AtaPassThru.h: Update PortMultiplierPort related comments
The UEFI2.5 Errata (Mantis ticket 1472 -
https://mantis.uefi.org/mantis/view.php?id=1472) updated
'PortMultiplierPort' parameter description for the following services in
EFI_ATA_PASS_THRU_PROTOCOL:
1. PassThru
2. BuildDevicePath
3. ResetDevice
Now, if there is no port multiplier on a ATA controller port, the
PortMultiplierPort parameter should be set to 0xFFFF instead of 0.
Please note that this is an incompatible semantic change. The consumer of
EFI_ATA_PASS_THRU_PROTOCOL needs to re-examine its usage to follow UEFI
2.5 Errata mantis 1472.
MdePkg Protocol/DevicePath.h: Update SATA Device Path comments
The UEFI2.5 spec (Mantis ticket 1353 -
https://mantis.uefi.org/mantis/view.php?id=1353) updated the Port
Multiplier Port Number description of SATA device path to use 0xFFFF when
the device is directly connected to the HBA.
Please note that this is an incompatible semantic change. The consumer of
SATA device path needs to re-examine its usage to follow UEFI 2.5 mantis
1353.
Eric Dong [Wed, 11 May 2016 02:36:05 +0000 (10:36 +0800)]
SecurityPkg OpalPasswordDxe: Error handling enhance when input password.
Enhance the error handling:
1. When the device is unlocked at BIOS phase and system does a warm reboot,
the device may be still in unlock status if it uses external power. For
such case, we would still popup password window to ask user input. If
user presses ESC key here, we would force the system shut down or ask
user input again to avoid security hole.
2. When user reach max retry count, force shutdown.
Zhang, Lubo [Wed, 11 May 2016 01:37:39 +0000 (09:37 +0800)]
NetworkPkg: Bug fix of iSCSI to support MPIO
If two attempts added on different NIC and enable
MPIO attribute, then change the attempts order. If
both two attempts succeed to connect the target,it
should abort the later one in the order and uninstall
ExtScsiPassThruProtocol Interface, But now it
unistall it twice.
Cc: Ye Ting <ting.ye@intel.com> Cc: Fu Siyuan <siyuan.fu@intel.com> Cc: Wu Jiaxin <jiaxin.wu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Zhang Lubo <lubo.zhang@intel.com> Reviewed-By: Ye Ting <ting.ye@intel.com> Reviewed-By: Fu Siyuan <siyuan.fu@intel.com> Reviewed-By: Wu Jiaxin <jiaxin.wu@intel.com>
ArmPlatformPkg/PL031RealTimeClockLib: remove local copy of gRT pointer
Since the only reason for keeping a local copy mRT of the gRT pointer
is to be able to call GetVariable/SetVariable at runtime, use the
UefiRuntimeLib helpers instead, so that we can drop mRT altogether.
PL031RealTimeClockLib is a base library that could potentially (although
unlikely) be incorporated into other modules than the DXE_RUNTIME_DRIVER
module that it was intended to complement.
This means the library has no business whatsoever setting the Runtime
Service table pointers directly (since we have no way of knowing which
instance will 'win', and the pointers may end up referring to a module
that is not a DXE_RUNTIME_DRIVER). So remove the assignment altogether.
Ryan Harkin [Wed, 11 May 2016 08:23:59 +0000 (09:23 +0100)]
EmbeddedPkg/Lan9118Dxe: add mask PCD to disable auto-negotiation features
Add a PCD to allow the platform to mask in/out specific features of
the LAN9118 device advertised during auto-negotiation.
For example, the Juno ARM Development Platform doesn't support full
duplex mode. This PCD will allow the platform developer to prevent the
full duplex modes from being advertised.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ryan Harkin <ryan.harkin@linaro.org>
[ardb: change default feature mask so that full duplex is disabled]
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Acked-by: Leif Lindholm <leif.lindholm@linaro.org>
Change PciHostBridgeDxe driver to not install the
PciHostBridgeResourceAllocation protocol and let
PciRootBridgeIo.Configuration() return the correct PCI resource
assignment information when the ResourceAssigned is TRUE.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com> Cc: Jeff Fan <jeff.fan@intel.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Ruiyu Ni [Fri, 29 Apr 2016 08:43:56 +0000 (16:43 +0800)]
MdeModulePkg/PciHostBridgeLib: Add ResourceAssigned field
Some platform doesn't require PciBus driver to assign resource
to PCI devices which causes PciRootBridgeIo.Configuration() cannot
return correct resource information to caller.
When resource assignment by PciBus is not mandatory, only light
version of PCI bus enumeration is performed, which only collects
the device resource consumption and publishes the PciIo instance.
The corresponding logic is in PciEnumeratorLight() in PciBus driver.
But PciEnumeratorLight() still depends on
PciRootBridgeIo.Configuration() returns the starting bus number in
order to search down to find all PCI devices under the root bridge.
When ResourceAssigned in PCI_ROOT_BRIDGE returned by
PciHostBridgeGetRootBridges() is TRUE, PciHostBridge driver treats
the Bus/Io/Mem/MemAbove4G/PMem/PMemAbove4G as the resource that are
*actually assigned* to the root bridge, instead of the resource that
*can be assigned* to the root bridge.
So that PciRootBridgeIo.Configuration() can return the correct
information.
Ruiyu Ni [Fri, 29 Apr 2016 09:38:51 +0000 (17:38 +0800)]
MdeModulePkg/PciHostBridgeDxe: Fix a Base/Limit comparing bug
When the aperture base equals to aperture limit, the old code treats
the aperture as non-existent. It's not correct because it indicates
a range starting with base and the length is 1.
The new code corrects the comparing bug.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com> Cc: Jeff Fan <jeff.fan@intel.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Linn Crosetto [Tue, 10 May 2016 18:16:12 +0000 (12:16 -0600)]
ArmVirtPkg: set PcdMaxVariableSize and PcdMaxAuthVariableSize
To support UEFI Secure Boot and the Linux persistent store with UEFI
variables, set PcdMaxVariableSize to 0x2000 bytes as is done in OvmfPkg.
For reference, the related Ovmf commits: 8cee3de72d441ca9
Also increase the maximum size for Authenticated variables in order to
handle a larger Signature List size as is done in OvmfPkg. Related Ovmf
commit: f5404a3e
Leahy, Leroy P [Mon, 9 May 2016 17:57:08 +0000 (10:57 -0700)]
CorebootModulePkg/BaseSerialPortLib: Set DTR and RTS
Ensure communication between the host and the UEFI system running
CorebootPayloadPkg. In cases where the host has flow control enabled
and the serial connection is providing the flow control signals, the
host will not be able to send data to the UEFI system because DTR and
RTS are not present. The host may also discard all output data from
the UEFI system because DTR is not present. By setting DTR and RTS
in the UART initialization code this case works properly.
Change-Id: I393f57104d111472cafcae01d4e43d4ea837be3b
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Lee Leahy <leroy.p.leahy@intel.com> Reviewed-by: Prince Agyeman <prince.agyeman@intel.com>
Mark Rutland [Fri, 6 May 2016 17:19:07 +0000 (18:19 +0100)]
EmbeddedPkg/Lan9118Dxe: add LAN9118 MMIO wrappers
As described in the LAN9118 datasheet, delays are necessary after some
reads and writes in order to ensure subsequent reads do not see stale
data.
This patch adds helpers to provide these delays automatically, by
performing dummy reads of the BYTE_TEST register (as recommended in the
LAN9118 datasheet). This approach allows the device register file itself
to provide the required delay, avoiding issues with early write
acknowledgement, or re-ordering of MMIO accesses aganist other
instructions (e.g. the delay loop).
Mark Rutland [Fri, 6 May 2016 17:19:06 +0000 (18:19 +0100)]
Revert "EmbeddedPkg/Lan9118Dxe: use MemoryFence"
Commit a4626006bbf86113 ("EmbeddedPkg/Lan9118Dxe: use MemoryFence")
replaced some stalls with memory fences, on the presumption that these
were erroneously being used to order memory accesses. However, this was
not the case.
LAN9118 devices require a timing delay between state-changing
reads/writes and subsequent reads, as updates to the register file are
asynchronous and the effects of state-changes are not immediately
visible to subsequent reads.
This delay cannot be ensured through the use of memory barriers, which
only enforce observable ordering, and not timing. Thus, converting these
stalls to memory fences was erroneous, and may result in stale values
being read.
ArmPkg/AArch64Mmu: don't let table entries inherit XN permission bits
When we split a block entry into a table entry, the UXN/PXN/XN permission
attributes are inherited both by the new table entry and by the new block
entries at the next level down. Unlike the NS bit, which only affects the
next level of lookup, the XN table bits supersede the permissions of the
final translation, and setting the permissions at multiple levels is not
only redundant, it also prevents us from lifting XN restrictions on a
subregion of the original block entry by simply clearing the appropriate
bits at the lowest level.
So drop the code that sets the UXN/PXN/XN bits on the table entries.
ArmPkg/ArmDmaLib: assert that consistent mappings are uncached
DmaMap () only allows uncached mappings to be used for creating consistent
mappings with operation type MapOperationBusMasterCommonBuffer. However,
if the buffer passed to DmaMap () happens to be aligned to the CWG, there
is no need for a bounce buffer, and we perform the cache maintenance
directly without ever checking if the memory attributes of the buffer
adhere to the API.
So add some debug code that asserts that the operation type and the memory
attributes are consistent.
ArmPkg/ArmDmaLib: do not remap arbitrary memory regions as uncached
In the DmaMap () operation, if the region to be mapped happens to be
aligned to the Cache Writeback Granule (CWG) (whose value is typically
64 or 128 bytes and 2 KB maximum), we remap the memory as uncached.
Since remapping memory occurs at page granularity, while the buffer and the
CWG may be much smaller, there is no telling what other memory we affect
by doing this, especially since the operation is not reverted in DmaUnmap().
ArmPkg/ArmDmaLib: reject consistent DMA mappings of cached memory
DmaMap () operations of type MapOperationBusMasterCommonBuffer should
return a mapping that is coherent between the CPU and the device. For
this reason, the API only allows DmaMap () to be called with this operation
type if the memory to be mapped was allocated by DmaAllocateBuffer (),
which in this implementation guarantees the coherency by using uncached
mappings on the CPU side.
This means that, if we encounter a cached mapping in DmaMap () with this
operation type, the code is either broken, or someone is violating the
API, but simply proceeding with a double buffer makes no sense at all,
and can only cause problems.
So instead, actively reject this operation type for cached memory mappings.
ArmPkg/ArmDmaLib: interpret GCD attributes as a bit field
Comparing a GCD attribute field directly against EFI_MEMORY_UC and
EFI_MEMORY_WT is incorrect, since it may have other bits set as well
which are not related to the cacheability of the region. So instead,
test explicitly against the flags EFI_MEMORY_WB and EFI_MEMORY_WT,
which must be set if the region may be mapped with cacheable attributes.
ArmPkg/ArmDmaLib: consistently use 'gCacheAlignment - 1' as alignment mask
We manage to use both an AND operation with 'gCacheAlignment - 1' and a
modulo operation with 'gCacheAlignment' in the same compound if statement.
Since gCacheAlignment is a global of which the compiler cannot guarantee
that it is a power of two, simply use the AND version in both cases.
ArmPkg/ArmDmaLib: deal with NULL return value of UncachedAllocatePages ()
The allocation function UncachedAllocatePages () may return NULL, in
which case our implementation of DmaAllocateBuffer () should return
EFI_OUT_OF_RESOURCES rather than silently ignoring the NULL value and
returning EFI_SUCCESS.
Ruiyu Ni [Mon, 9 May 2016 03:21:25 +0000 (11:21 +0800)]
MdeModulePkg/PciSioSerialDxe: Do not flush the UART
The patch aligns to the IntelFrameworkModulePkg/Bus/Isa/IsaSerialDxe
driver not flush the UART in Reset() and SetAttributes() function.
It was found the flush causes hang on certain PCI serial devices.
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com> Reviewed-by: Eric Jin <eric.jin@intel.com>
MdeModulePkg RamDiskDxe: Fix wrong HII behavior for more than 8 RAM disks
The RamDiskDxe driver originally uses a variable-length HII varstore to
retrieve the HII checkbox status of each registered RAM disk.
However, HII does not support the variable-length varstore feature.
Therefore, only the checkbox status for the first 8 RAM disks are tracked
for the following definition of HII varstore structure considering the
alignment:
OvmfPkg: Modify FDF/DSC files for RamDiskDxe's adding NFIT report feature
The RamDiskDxe driver in MdeModulePkg now will use EFI_ACPI_TABLE_PROTOCOL
and EFI_ACPI_SDT_PROTOCOL during reporting RAM disks to NVDIMM Firmware
Interface Table (NFIT).
A Pcd 'PcdInstallAcpiSdtProtocol' controls whether the
EFI_ACPI_SDT_PROTOCOL will be produced. Its default value is set to FALSE
in MdeModulePkg. To make the NFIT reporting feature working properly under
OVMF, the patch will set the Pcd to TRUE in OVMF DSC files.
Also, the RamDiskDxe driver will sometimes report a NVDIMM Root Device
using ASL code which is put in a Secondary System Description Table (SSDT)
according to the ACPI 6.1 spec.
Locating the SSDT requires modifying the [Rule.Common.DXE_DRIVER] field in
OVMF FDF files.
Ard Biesheuvel [Mon, 9 May 2016 14:24:18 +0000 (16:24 +0200)]
ArmPkg/DefaultExceptionHandlerLib: add stack dump to exception handling code
This adds a partial stack dump (256 bytes at either side of the stack
pointer) to the CPU state dumping routine that is invoked when taking an
unexpected exception. Since dereferencing the stack pointer may itself
fault, ensure that we don't enter the dumping routine recursively.
Ard Biesheuvel [Mon, 9 May 2016 14:14:15 +0000 (16:14 +0200)]
ArmPkg/DefaultExceptionHandlerLib: use deadloop rather than ASSERT
The default exception handler, which is essentially the one that is invoked
for unexpected exceptions, ends with an ASSERT (FALSE), to ensure that
execution halts after dumping the CPU state. However, ASSERTs are compiled
out in RELEASE builds, and since we simply return to wherever the ELR is
pointing, we will not make any progress in case of synchronous aborts, and
the same exception will be taken again immediately, resulting in the string
'Exception at 0x....' to be printed over and over again.
MdeModulePkg/SdMmcPciHcDxe: Using PIO rather than DMA for clock tuning
The original code is using ADMA mode to do clock tuning procedure. It
may have problem on some SD/MMC host controllers as there is no way to
know when to send next tuning cmd.
Update it to PIO mode to strictly follow SD Host Controller Simplified
Specification 3.0 Figure 2-29. By this way, if the Buffer Read Ready
interrupt is set, we could know it's ok to send the next clock tuning
cmd.
MdeModulePkg/SdMmcPciHcDxe: Use BaseClk if the target clock is larger
The original code has a bug to calculate which clock freq should be
used when the target clock freq is larger than the BaseClock Freq
provided by the system.
Unlike in OVMF, no USE_OLD_BDS fallback is introduced; I think that
ArmVirtPkg is less widely used by non-developers than OvmfPkg.
ArmVirtXen is not modified, as it uses PlatformIntelBdsLib from
ArmPlatformPkg.
About this patch:
- DxeServicesLib and SortLib are resolved generally (they have broad
client module type lists).
- ReportStatusCodeLib is resolved for UEFI_APPLICATION modules.
- GenericBdsLib and PlatformBdsLib are replaced with UefiBootManagerLib
and PlatformBootManagerLib, and resolved from under MdeModulePkg and
ArmVirtPkg, respectively.
- QemuBootOrderLib is pointed to the QemuNewBootOrderLib instance.
- FileExplorerLib no longer depends on SECURE_BOOT_ENABLE, it is nedeed by
BootMaintenanceManagerUiLib, which we link into UiApp.
- PcdBootManagerMenuFile carries the FILE_GUID of
"MdeModulePkg/Application/UiApp/UiApp.inf". The default PCD value from
"MdeModulePkg/MdeModulePkg.dec" points to
"MdeModulePkg/Application/BootManagerMenuApp/BootManagerMenuApp.inf",
which, according to the commit that introduced it (a382952f8255), only
'provides a very simple UI showing all the boot options recorded by
"BootOrder" and user can select any of them to boot'.
- Include the new core BDS driver, and include the boot manager
application, with the usual main menu entries.
UefiBootManagerLib does not provide these functions, we have to implement
them. (EnableQuietBoot() puts up the nice TianoCore logo.)
OvmfPkg commits 817fb3ac2a78 and 8e8fd3037788 have extracted these
functions already,
- from "IntelFrameworkModulePkg/Library/GenericBdsLib/BdsConsole.c"
- to "OvmfPkg/Library/PlatformBootManagerLib/QuietBoot.c".
Laszlo Ersek [Thu, 5 May 2016 15:45:42 +0000 (17:45 +0200)]
ArmVirtPkg/PlatformBootManagerLib: rebase boot order manipulation
QemuBootOrderLib can only filter out and reorder boot options; it cannot
create boot options. It relies on Platform BDS to auto-generate all
possible boot options first (for example, for new virtual devices that
have been configured since the last run of the virtual machine). Then it
will decide, case-by-case, whether each of those auto-generated boot
options should be preserved (and at what position), or removed.
Thus far, the only implementation of SetBootOrderFromQemu(), used in
connection with IntelFrameworkModulePkg BDS, has expected said complete
boot option list as an input parameter:
BdsEntry() [IntelFrameworkModulePkg/Universal/BdsDxe/BdsEntry.c]
// create empty list
InitializeListHead (
&BootOptionList
)
// The "BootOrder" variable may have changed. Refresh BootOptionList
// from it, and return it to BdsEntry().
With MdeModulePkg BDS, a BootOptionList is not propagated from BdsEntry()
to SetBootOrderFromQemu() and back. All processing is based directly on
the underlying "BootOrder" and "Boot####" variables.
In OvmfPkg, commit d27ec22d1189 introduced a new instance of
QemuBootOrderLib, called QemuNewBootOrderLib. It is based on
UefiBootManagerLib, and rather than taking a complete BootOptionList as a
parameter, it expects that the "BootOrder" and "Boot####" variables are
complete in the above sense.
Rebase the boot order manipulation to UefiBootManagerLib and
QemuNewBootOrderLib, while keeping the requirement satisfied, like this:
Laszlo Ersek [Thu, 5 May 2016 15:22:33 +0000 (17:22 +0200)]
ArmVirtPkg/PlatformBootManagerLib: register boot options and hot keys
Register the Enter key as the continue key (hot key to skip the boot
timeout). Map the F2 and ESC keys to the UI. Register the memory-mapped
Shell boot option.
The patch parallels OvmfPkg commit 07dd96e82032. The
PlatformRegisterFvBootOption() and PlatformRegisterOptionsAndKeys()
functions are copied almost verbatim. The only changes are: internal
linkage for these functions (i.e., STATIC), and mentioning the ESC key in
the comments.