Anthony PERARD [Wed, 31 Jan 2018 16:01:26 +0000 (16:01 +0000)]
OvmfPkg/OvmfXen: Introduce XenDebugLibIoPort
Introduce XenDebugLibIoPort which is enabled with
DEBUG_ON_HYPERVISOR_CONSOLE which send the debug output to Xen's
console.
It's a copy PlatformDebugLibIoPort which always write to the IO port.
It works with both Xen HVM guest and Xen PVH guest whereas the default
PlatformDebugLibIoPort works only in HVM when QEMU is present.
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Anthony PERARD [Mon, 20 Apr 2020 12:45:15 +0000 (13:45 +0100)]
OvmfPkg/OvmfXen: Remove DEBUG_ON_SERIAL_PORT
Remove support for DEBUG_ON_SERIAL_PORT because OvmfXen can't be build
with it due to a circular dependency:
DebugLib : BaseDebugLibSerialPort ->
SerialPortLib : XenConsoleSerialPortLib ->
XenHypercallLib : XenHypercallLib ->
DebugLib
Also, if that dependency is fixed, I think it would be harder to find
which console the debug is sent to when running an HVM guest. The xen
console isn't the serial console used by default. Furthermore,
XenHypercallLib isn't initialised early enough, so we would loose
debug output from the SEC phase and early PEI phase.
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
According to EDK2 Driver Writer's Guide For UEFI 2.3.1, 4.2.18 Offer
alternatives to function keys. Configuration of drivers should be
accomplished via HII and via OS-present interfaces.
Cc: Jiaxin Wu <jiaxin.wu@intel.com> Cc: Siyuan Fu <siyuan.fu@intel.com> Signed-off-by: Maciej Rabeda <maciej.rabeda@linux.intel.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com> Reviewed-by: Maciej Rabeda <maciej.rabeda@linux.intel.com>
Ip6ConfigReadConfigData() reads configuration data from a UEFI variable
and copies the data to another buffer. This change checks that the
length
of the data record being copied does not exceed the size of the source
UEFI variable data buffer.
If the size is exceeded, this change follows existing logic to treat the
variable as corrupted and deletes the variable so it will be set again.
Cc: Siyuan Fu <siyuan.fu@intel.com> Cc: Maciej Rabeda <maciej.rabeda@linux.intel.com> Cc: Jiaxin Wu <jiaxin.wu@intel.com> Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com> Reviewed-by: Maciej Rabeda <maciej.rabeda@linux.intel.com>
SnpDxe initializes values for MemoryBarIndex and IoBarIndex to 0 and 1
respectively even if calls to PciIo->GetBarAttributes never return
success.
Later, if the BAR is used to perform IO/Mem reads/writes, a potentially
non-existent BAR index may be accessed. This change initializes the
values
to an invalid BAR index (PCI_MAX_BAR) so the condition can be explicitly
checked to avoid an invalid BAR access.
Cc: Siyuan Fu <siyuan.fu@intel.com> Cc: Maciej Rabeda <maciej.rabeda@linux.intel.com> Cc: Jiaxin Wu <jiaxin.wu@intel.com> Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com> Reviewed-by: Maciej Rabeda <maciej.rabeda@linux.intel.com>
The current SnpDxe implementation registers its ExitBootServices event
notification function (SnpNotifyExitBootServices ()) at TPL_NOTIFY. This
function calls PxeShutdown() which issues an UNDI shutdown operation.
Ultimately, this may invoke Shutdown() in EFI_SIMPLE_NETWORK_PROTOCOL.
The UEFI specification 2.8A Table 27 "TPL Restrictions" restricts the
TPL
for Simple Network Protocol to <= TPL_CALLBACK. In addition, it has been
observed in some 3rd party UNDI drivers to cause an issue further down
the call stack if the TPL is higher than TPL_CALLBACK on invocation.
Therefore, this commit changes the TPL of SnpNotifyExitBootServices() to
TPL_CALLBACK.
Cc: Siyuan Fu <siyuan.fu@intel.com> Cc: Maciej Rabeda <maciej.rabeda@linux.intel.com> Cc: Jiaxin Wu <jiaxin.wu@intel.com> Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Maciej Rabeda <maciej.rabeda@linux.intel.com>
Pierre Gondois [Mon, 30 Mar 2020 08:52:38 +0000 (16:52 +0800)]
MdePkg: Add AML FieldList OpCode definitions
The ACPI specification, version 6.3, January 2019,
defines the Named Objects Encoding for FieldElements
in section '20.2.5.2 Named Objects Encoding'.
FieldElements can be one of the following:
NamedField | ReservedField | AccessField |
ExtendedAccessField | ConnectField
Some of these keywords are starting with an opcode,
allowing to identify their type.
E.g.: ReservedField := 0x00 PkgLength
This patch adds these FieldElement opcodes definitions to
the list of AML Opcode definitions.
Reviewed-by: Liming Gao <liming.gao@intel.com> Reviewed-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Zhiguang Liu <zhiguang.liu@intel.com> Signed-off-by: Pierre Gondois <pierre.gondois@arm.com>
MdeModulePkg/NonDiscoverablePciDeviceDxe: use standard PCI_MAX_BAR macro
NonDiscoverablePciDeviceDxe currently #defines the module-internal
PCI_MAX_BARS macro, with value 6 (and type "int").
Remove this private macro definition, and adopt PCI_MAX_BAR from
"MdePkg/Include/IndustryStandard/Pci22.h". Both the value and the type
remain unchanged (the standard PCI_MAX_BAR macro expands to 0x0006).
"NonDiscoverablePciDeviceIo.h" gets the definition of PCI_MAX_BAR via its
existent #include <IndustryStandard/Pci.h> directive.
Build-tested only.
Cc: Ard Biesheuvel <ard.biesheuvel@arm.com> Cc: Hao A Wu <hao.a.wu@intel.com> Cc: Jian J Wang <jian.j.wang@intel.com> Cc: Philippe Mathieu-Daudé <philmd@redhat.com> Cc: Ray Ni <ray.ni@intel.com> Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20200409113017.18233-1-lersek@redhat.com> Reviewed-by: Ard Biesheuvel <ard.biesheuvel@arm.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Ray Ni <ray.ni@intel.com>
Shenglei Zhang [Fri, 21 Feb 2020 02:21:06 +0000 (10:21 +0800)]
BaseTools: Make brotli a submodule
Use submodule way to access brotli in BaseTools based on
brotli version 666c3280cc11dc433c303d79a83d4ffbdd12cc8d.
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2558
Shenglei Zhang [Wed, 12 Feb 2020 10:08:15 +0000 (18:08 +0800)]
MdeModulePkg/BrotliCustomDecompressLib: Make brotli a submodule
Use submodule way to access brotli in MdeModulePkg based on
brotli version 666c3280cc11dc433c303d79a83d4ffbdd12cc8d.
The newly added BrotliDecUefiSupport.h/.c are used by directory
'brotli'.
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2559
MdePkg/UefiScsiLib: Set FUA bit for synchronous SCSI Write operations
The FUA (Force Unit Access) bit forces data to be written directly to
disk instead of the write cache. This prevents data from being lost if a
shutdown or reset is requested immediately after a SCSI write operation.
Cc: Michael D Kinney <michael.d.kinney@intel.com> Cc: Jian J Wang <jian.j.wang@intel.com> Cc: Liming Gao <liming.gao@intel.com> Cc: Zhiguang Liu <zhiguang.liu@intel.com> Reviewed-by: Liming Gao <liming.gao@intel.com> Reviewed-by: Zhiguang Liu <zhiguang.liu@intel.com> Signed-off-by: Christopher J Zurcher <christopher.j.zurcher@intel.com>
Zhiguang Liu [Wed, 18 Mar 2020 01:39:30 +0000 (09:39 +0800)]
MdePkg DebugLib: Enable FILE NAME as DEBUG ASSERT for CLANG
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1548
For clang compiler, This change will make the generated image not include the
absolute file path. If so, the generated debug image can be reproduced in the
different root directory. Also, it will reduce the size of debug image size.
To do so, use __FILE_NAME__ macro to replace __FILE__. __FILE_NAME__ is
Clang-specific extension that functions similar to __FILE__ but only renders
the last path component (the filename) instead of an invocation dependent full
path to that file, and __FILE_NAME__ is introduced since clang 9.
Cc: Michael D Kinney <michael.d.kinney@intel.com> Cc: Liming Gao <liming.gao@intel.com> Reviewed-by: Liming Gao <liming.gao@intel.com> Signed-off-by: Zhiguang Liu <zhiguang.liu@intel.com>
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2659
Allow WindowsVsToolChain Plugin to add libraries and headers
of user defined ARCH for VS2017 and VS2019.
pip-requirements.txt: Update extensions min version to 0.13.3
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2616
Pytool extensions are locked on 0.12.x but extensions has
moved to 0.13.x. So update the pip-requirements.txt.
Cc: Sean Brogan <sean.brogan@microsoft.com> Cc: Bret Barkelew <Bret.Barkelew@microsoft.com> Cc: Michael D Kinney <michael.d.kinney@intel.com> Cc: Liming Gao <liming.gao@intel.com> Signed-off-by: Shenglei Zhang <shenglei.zhang@intel.com> Reviewed-by: Bret Barkelew <bret.barkelew@microsoft.com>
After remove Used parameter, below code in ResetTokens can also be
removed:
1. The RunningApCount parameter will be reset in GetFreeToken.
2. The ReleaseSpinLock should be called in ReleaseToken function,
Code in this function seems like a later fix if ReleaseToken not
Release it. We should remove code here and fix the real issue if
existed.
Signed-off-by: Eric Dong <eric.dong@intel.com> Reviewed-by: Ray Ni <ray.ni@intel.com> Cc: Star Zeng <star.zeng@intel.com> Cc: Laszlo Ersek <lersek@redhat.com>
After patch "UefiCpuPkg/PiSmmCpuDxeSmm: Improve the
performance of GetFreeToken()" which adds new parameter
FirstFreeToken, it's not need to use Uses parameter.
This patch used to remove this parameter.
Signed-off-by: Eric Dong <eric.dong@intel.com> Reviewed-by: Ray Ni <ray.ni@intel.com> Cc: Star Zeng <star.zeng@intel.com> Cc: Laszlo Ersek <lersek@redhat.com>
Ray Ni [Fri, 10 Apr 2020 02:51:26 +0000 (10:51 +0800)]
UefiCpuPkg/PiSmmCpuDxeSmm: Improve the performance of GetFreeToken()
Today's GetFreeToken() runs at the algorithm complexity of O(n)
where n is the size of the token list.
The change introduces a new global variable FirstFreeToken and it
always points to the first free token. So the algorithm complexity
of GetFreeToken() decreases from O(n) to O(1).
The improvement matters when some SMI code uses StartupThisAP()
service for each of the AP such that the algorithm complexity
becomes O(n) * O(m) where m is the AP count.
As next steps,
1. PROCEDURE_TOKEN.Used field can be optimized out because
all tokens before FirstFreeToken should have "Used" set while all
after FirstFreeToken should have "Used" cleared.
2. ResetTokens() can be optimized to only reset tokens before
FirstFreeToken.
v2: add missing line in InitializeDataForMmMp.
v3: update copyright year to 2020.
Signed-off-by: Ray Ni <ray.ni@intel.com> Reviewed-by: Eric Dong <eric.dong@intel.com> Reviewed-by: Star Zeng <star.zeng@intel.com>
Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Andrew Fish <afish@apple.com> Cc: Ray Ni <ray.ni@intel.com> Signed-off-by: Guomin Jiang <guomin.jiang@intel.com> Reviewed-by: Ray Ni <ray.ni@intel.com>
The link flags for VS2019 is absent and result the build fail.
Cc: Jordan Justen <jordan.l.justen@intel.com> Cc: Andrew Fish <afish@apple.com> Cc: Ray Ni <ray.ni@intel.com> Signed-off-by: Guomin Jiang <guomin.jiang@intel.com> Reviewed-by: Ray Ni <ray.ni@intel.com>
Shenglei Zhang [Fri, 27 Dec 2019 06:28:20 +0000 (14:28 +0800)]
MdeModulePkg/RegularExpressionDxe: Make oniguruma a submodule in edk2.
Use submodule way to access oniguruma. And upgrade oniguruma
version from v6.9.3 to v6.9.4_mark1.
REF: https://bugzilla.tianocore.org/show_bug.cgi?id=2073
This patch reduces the size of IA32 binaries by ensuring that
no .debug_frame / .eh_frame sections are generated through forcing
SEH exception model, which is already the default in clang for X64.
EDK II does not support exceptions, and in future we should disable
them instead of switching to some other variant. Currently this
is not possible due to the following LLVM bugs:
Index should be off-by one than size of array, so when check
array, the max index should less than size of array.
Cc: Jian J Wang <jian.j.wang@intel.com> Cc: Hao A Wu <hao.a.wu@intel.com> Cc: Eric Dong <eric.dong@intel.com> Cc: Ray Ni <ray.ni@intel.com> Signed-off-by: Guomin Jiang <guomin.jiang@intel.com> Reviewed-by: Jian J Wang <jian.j.wang@intel.com>
According to CryptoPkg.dsc, the library class only have HashApiLib, so i
think the BaseHashApiLib should be considered as base name rather than
library class.
Cc: Jian J Wang <jian.j.wang@intel.com> Cc: Xiaoyu Lu <xiaoyux.lu@intel.com> Signed-off-by: Guomin Jiang <guomin.jiang@intel.com> Reviewed-by: Shenglei Zhang <shenglei.zhang@intel.com> Reviewed-by: Liming Gao <liming.gao@intel.com> Reviewed-by: Xiaoyu Lu <xiaoyux.lu@intel.com>
GetVariable() now returns attributes when it fails with
EFI_BUFFER_TOO_SMALL. Therefore, commit 6896efdec270 is
reverted since it is no longer relevant.
Cc: Laszlo Ersek <lersek@redhat.com> Cc: Siyuan Fu <siyuan.fu@intel.com> Cc: Maciej Rabeda <maciej.rabeda@linux.intel.com> Cc: Jiaxin Wu <jiaxin.wu@intel.com> Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com> Reviewed-by: Bret Barkelew <bret.barkelew@microsoft.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com>
This commit makes the behavior for PeiGetVariable() match the following
specification-defined behavior. It is now consistent with the DXE/SMM
variable driver implementation.
The UEFI specification v2.8 Errata A Section 8.2 "GetVariable()"
"Attributes" parameter description states:
"If not NULL, a pointer to the memory location to return the
attributes bitmask for the variable. See 'Related Definitions.'
If not NULL, then Attributes is set on output both when
EFI_SUCCESS and when EFI_BUFFER_TOO_SMALL is returned."
The attributes were previously only returned from the implementation
in Variable.c on EFI_SUCCESS. They are now returned on EFI_SUCCESS or
EFI_BUFFER_TOO_SMALL according to spec.
Cc: Bret Barkelew <Bret.Barkelew@microsoft.com> Cc: Liming Gao <liming.gao@intel.com> Cc: Michael D Kinney <michael.d.kinney@intel.com> Cc: Jian J Wang <jian.j.wang@intel.com> Cc: Hao A Wu <hao.a.wu@intel.com> Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com> Reviewed-by: Jian J Wang <jian.j.wang@intel.com> Reviewed-by: Guomin Jiang <guomin.jiang@intel.com>
The UEFI specification v2.8 Errata A Section 8.2 "GetVariable()"
"Attributes" parameter description states:
"If not NULL, a pointer to the memory location to return the
attributes bitmask for the variable. See 'Related Definitions.'
If not NULL, then Attributes is set on output both when
EFI_SUCCESS and when EFI_BUFFER_TOO_SMALL is returned."
The attributes were previously only returned from the implementation
in Variable.c on EFI_SUCCESS. They are now returned on EFI_SUCCESS or
EFI_BUFFER_TOO_SMALL according to spec.
Cc: Bret Barkelew <Bret.Barkelew@microsoft.com> Cc: Liming Gao <liming.gao@intel.com> Cc: Michael D Kinney <michael.d.kinney@intel.com> Cc: Jian J Wang <jian.j.wang@intel.com> Cc: Hao A Wu <hao.a.wu@intel.com> Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com> Reviewed-by: Guomin Jiang <guomin.jiang@intel.com> Reviewed-by: Jian J Wang <jian.j.wang@intel.com>
MdeModulePkg/DxeCore: merge properties table routines into MAT handling
The memory attributes table handling code shared some parts of the
properties table handling code when it was introduced. The latter has
been dropped, and so we can merge the remaining pieces into the former.
Link: https://bugzilla.tianocore.org/show_bug.cgi?id=2633 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@arm.com> Reviewed-by: Bret Barkelew <bret.barkelew@microsoft.com> Acked-by: Jiewen Yao <Jiewen.yao@intel.com> Acked-by: Liming Gao <liming.gao@intel.com> Reviewed-by: Dandan Bi <dandan.bi@intel.com> Reviewed-by: Jian J Wang <jian.j.wang@intel.com>
MdeModulePkg: disable properties table generation but retain the code
This is the minimal change required to stop exposing the EFI properties
table, which is deprecated. Given how the implementation is entangled
with the code that exposes the related memory attributes table, most of
the code is retained, and further cleanups are relegated to subsequent
patches.
Link: https://bugzilla.tianocore.org/show_bug.cgi?id=2633 Signed-off-by: Ard Biesheuvel <ard.biesheuvel@arm.com> Acked-by: Jiewen Yao <Jiewen.yao@intel.com> Acked-by: Liming Gao <liming.gao@intel.com> Reviewed-by: Dandan Bi <dandan.bi@intel.com> Reviewed-by: Jian J Wang <jian.j.wang@intel.com>
The UEFI properties table and the associated memory protection feature was
severely broken from the start, and has been deprecated for a while. Let's
drop all references to it from OVMF so we can safely remove it from the
DXE core as well.
Azure Pipelines agents have 2 threads. This commit has been shown to
reduce the build time in half on those agents.
Cc: Bob C Feng <bob.c.feng@intel.com> Cc: Liming Gao <liming.gao@intel.com> Signed-off-by: Michael Kubacki <michael.kubacki@microsoft.com> Reviewed-by: Bob C Feng <bob.c.feng@intel.com>
The local #define TT_ATTR_INDX_INVALID is used as a local error code
in the AArch64 implementation, but is misleadingly named to match the
definitions in ArmPkg/Include/Chipset/AArch64Mmu.h.
Rename it INVALID_ENTRY to reduce confusion and improve readability.
Signed-off-by: Abner Chang <abner.chang@hpe.com> Reviewed-by: Ray Ni <ray.ni@intel.com> Cc: Ray Ni <ray.ni@intel.com> Cc: Leif Lindholm <leif@nuviainc.com> Cc: Gilbert Chen <gilbert.chen@hpe.com> Cc: Daniel Schaefer <daniel.schaefer@hpe.com>
Ard Biesheuvel [Tue, 31 Mar 2020 17:25:11 +0000 (19:25 +0200)]
ArmPkg/ArmMmuLib: drop unused TT_ATTR_INDX_INVALID CPP macro
TT_ATTR_INDX_INVALID is #define'd but never used so drop it. Note
that this leaves a CPP macro of the same name in CpuDxe, but there,
it is actually being used, and although the name suggests that this
value is somehow defined by the architecture, this is really not the
case and it only has meaning within the scope of CpuDxe's implementation.
Ard Biesheuvel [Tue, 31 Mar 2020 17:25:06 +0000 (19:25 +0200)]
ArmPkg/ArmMmuLib: get rid of GetRootTranslationTableInfo()
Only a single call to GetRootTranslationTableInfo() remains, which
only provides the root table level. So let's create a new static
helper function that returns just this value, and use it instead.
Ard Biesheuvel [Tue, 31 Mar 2020 17:25:02 +0000 (19:25 +0200)]
ArmPkg/ArmMmuLib: drop pointless LookupAddresstoRootTable() routine
LookupAddresstoRootTable() uses a loop to go over its MaxAddress
argument, essentially to do a log2() and determine how many bits are
needed to represent it. Since the argument is the result of a shift-left
expression, there is some room for improvement here, and we can simply
use the bit count directly to calculate the value of T0SZ. At the same
time, we can omit calling GetRootTranslationTableInfo() to determine the
number of root table entries, and add a new helper that applies the
trivial calculation directly.
Ard Biesheuvel [Tue, 31 Mar 2020 17:24:57 +0000 (19:24 +0200)]
ArmPkg/CpuDxe: move PageAttributeToGcdAttribute() out of ArmMmuLib
The routine PageAttributeToGcdAttribute() is exported by ArmMmuLib
but only ever used in the implementation of CpuDxe. So let's move
the function there and make it STATIC.
Ard Biesheuvel [Tue, 31 Mar 2020 17:24:45 +0000 (19:24 +0200)]
ArmPkg/CpuDxe: use private copy of GetRootTranslationTableInfo()
Before getting rid of GetRootTranslationTableInfo() and the related
LookupAddresstoRootTable() in AARCH64's version of ArmMmuLib, add a
version of the former to CpuDxe, which will be its only remaining
user. While at it, simplify it a bit, since in the CpuDxe cases,
both OUT arguments are always provided.
Note that this removes the declaration of GetRootTranslationTableInfo()
as well, but this is a declaration that is private to CpuDxe, and it
really doesn't belong here in the first place. Since ArmMmuLib's version
of GetRootTranslationTableInfo() is going to be replaced shortly anyway,
don't bother moving this .h declaration elsewhere.
This patch fixes reversed logic of recently added ASSERTs which should
ensure that Ip6IsNDOptionValid() implementation properly reacts to invalid
packets.
Cc: Jiaxin Wu <jiaxin.wu@intel.com> Cc: Siyuan Fu <siyuan.fu@intel.com> Signed-off-by: Maciej Rabeda <maciej.rabeda@linux.intel.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com> Tested-by: Laszlo Ersek <lersek@redhat.com> Fixes: 9c20342eed70ec99ec50cd73cb81804299f05403
Liran Alon [Tue, 31 Mar 2020 22:56:37 +0000 (01:56 +0300)]
OvmfPkg/PvScsiDxe: Refactor setup of rings to separate function
Previous to this change, PvScsiFreeRings() was not undoing all
operations that was done by PvScsiInitRings().
This is because PvScsiInitRings() was both preparing rings (Allocate
memory and map it for device DMA) and setup the rings against device by
issueing a device command. While PvScsiFreeRings() only unmaps the rings
and free their memory.
Driver do not have a functional error as it makes sure to reset device
before every call site to PvScsiFreeRings(). However, this is not
intuitive.
Therefore, prefer to refactor the setup of the ring against device to a
separate function than PvScsiInitRings().
Liran Alon [Tue, 31 Mar 2020 11:04:52 +0000 (14:04 +0300)]
OvmfPkg/PvScsiDxe: Fix VS2019 build error because of implicit cast
Sean reported that VS2019 build produce the following build error:
INFO - PvScsi.c
INFO - Generating code
INFO - d:\a\1\s\OvmfPkg\PvScsiDxe\PvScsi.c(459): error C2220: the
following warning is treated as an error
INFO - d:\a\1\s\OvmfPkg\PvScsiDxe\PvScsi.c(459): warning C4244: '=':
conversion from 'const UINT16' to 'UINT8', possible loss of data
This result from an implicit cast from PVSCSI Response->ScsiStatus
(Which is UINT16) to Packet->TargetResponse (Which is UINT8).
Fix this issue by adding an appropriate explicit cast and verify with
assert that this truncation do not result in loss of data.
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=2651 Reported-by: Sean Brogan <sean.brogan@microsoft.com> Signed-off-by: Liran Alon <liran.alon@oracle.com>
Message-Id: <20200331110452.51992-1-liran.alon@oracle.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com>
[lersek@redhat.com: rewrap VS2019 diags in commit msg for PatchCheck.py]
Liran Alon [Tue, 31 Mar 2020 11:02:44 +0000 (14:02 +0300)]
Maintainers.txt: Add Liran and Nikita as OvmfPkg/PvScsiDxe reviewers
Laszlo suggested that as I have contributed the OvmfPkg PVSCSI driver, I
will also register myself as a reviewer in Maintainers.txt.
In addition, as Nikita have assisted the development of the PVSCSI
driver and have developed another similar OvmfPkg SCSI driver, add him
as a reviewer to PVSCSI driver as-well.
When DHCP is misconfigured on a network segment, such that two DHCP
servers attempt to reply to requests (and therefore race with each other),
the edk2 PXE client can confuse itself.
In PxeBcDhcp4BootInfo() / PxeBcDhcp6BootInfo(), the client may refer to a
DHCP reply packet as an "earlier" packet from the "same" DHCP server, when
in reality both packets are unrelated, and arrive from different DHCP
servers.
While the edk2 PXE client can do nothing to fix this, it should at least
not ASSERT() -- ASSERT() is for catching programming errors (violations of
invariants that are under the control of the programmer). ASSERT()s should
in particular not refer to external data (such as network packets). What's
more, in RELEASE builds, we get NULL pointer references.
Check the problem conditions with actual "if"s, and return
EFI_PROTOCOL_ERROR. This will trickle out to PxeBcLoadBootFile(), and be
reported as "PXE-E99: Unexpected network error".
Cc: Jiaxin Wu <jiaxin.wu@intel.com> Cc: Maciej Rabeda <maciej.rabeda@linux.intel.com> Cc: Philippe Mathieu-Daudé <philmd@redhat.com> Cc: Siyuan Fu <siyuan.fu@intel.com> Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Message-Id: <20200331004749.16128-1-lersek@redhat.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com> Reviewed-by: Maciej Rabeda <maciej.rabeda@linux.intel.com>
'ping' command's receive flow utilizes a single Rx token which it
attempts to reuse before recycling the previously received packet.
This causes a situation where under ICMP traffic,
Ping6OnEchoReplyReceived() function will receive an already
recycled packet with EFI_SUCCESS token status and finally
dereference invalid pointers from RxData structure.
Cc: Ray Ni <ray.ni@intel.com> Cc: Zhichao Gao <zhichao.gao@intel.com> Signed-off-by: Maciej Rabeda <maciej.rabeda@linux.intel.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com> Acked-by: Zhichao Gao <zhichao.gao@intel.com>
On enabling the /analyse option the VS2017 compiler
reports: warning C6001: Using uninitialized memory.
This warning is reported for the Status variable in
AddGenericInitiatorAffinity() as it is not initialised
to a default value. This condition is only valid if
GenInitAffCount is equal to 0. Since GenInitAffCount
is already checked in BuildSratTable() this condition
can never happen.
The value of the Status variable is returned in
failure cases from appropriate locations in
AddGenericInitiatorAffinity(). The only case
where Status value is being used un-initialised
is the return statement at the end of
AddGenericInitiatorAffinity().
Therefore, to fix this issue EFI_SUCCESS can be
safely returned instead of returning the Status
variable at the end of the function.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com>
Sami Mujawar [Mon, 18 Nov 2019 16:30:16 +0000 (16:30 +0000)]
DynamicTablesPkg: SRAT: Fix entry points
VS2017 reports 'warning C4028: formal parameter 2 different
from declaration' for the library constructor and destructor
interfaces for the SRAT Generator modules.
Remove the CONST qualifier for the ImageHandle and the
SystemTable pointer in the library constructor and destructor
to make it compatible with the formal declaration.
Ard Biesheuvel [Thu, 5 Mar 2020 11:17:23 +0000 (12:17 +0100)]
Maintainers: switch to my Arm email address
I no longer work for Linaro (and haven't for a while) so in anticipation
of losing access to my @linaro.org mailbox, let's switch to the ARM one
for my Tiancore contributions and maintainerships.
Cc: Andrew Fish <afish@apple.com> Cc: Michael D Kinney <michael.d.kinney@intel.com> Cc: Ard Biesheuvel <ard.biesheuvel@arm.com> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@arm.com> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Reviewed-by: Leif Lindholm <leif@nuviainc.com> Reviewed-by: Laszlo Ersek <lersek@redhat.com>
For LzmaCompress or BrotliCompress, the platform may use the different
options and add their batch file, such as LzmaCompressPlatform.
Then, specify it in platform.dsc [BuildOptions] to override the default
one in tools_def.txt.
*_*_*_LZMA_PATH = LzmaCompressPlatform
This override tool will be used. But, its name is not specified in the
generated GuidedSectionTools.txt.
Signed-off-by: Zhiju.Fan <zhijux.fan@intel.com> Cc: Liming Gao <liming.gao@intel.com> Cc: Bob Feng <bob.c.feng@intel.com> Reviewed-by: Bob Feng <bob.c.feng@intel.com>
Liran Alon [Sat, 28 Mar 2020 20:00:59 +0000 (23:00 +0300)]
OvmfPkg/PvScsiDxe: Reset device on ExitBootServices()
This causes the device to forget about the request/completion rings.
We allocated said rings in EfiBootServicesData type memory, and code
executing after ExitBootServices() is permitted to overwrite it.
Liran Alon [Sat, 28 Mar 2020 20:00:57 +0000 (23:00 +0300)]
OvmfPkg/PvScsiDxe: Introduce DMA communication buffer
In case device is constrained by IOMMU or guest is running under AMD SEV,
input/output buffers provided to device (DataBuffer and SenseData) needs
to be explicitly mapped to device by PciIo->Map().
To avoid the overhead of mapping/unmapping the DataBuffer and SenseData
to the device for every SCSI requst (and to simplify code), introduce a
single DMA communication buffer that will be mapped to device on
initialization. When a SCSI request needs to be sent to device, the
DataBuffer and SenseData will be copied from/to the DMA communication
buffer as required. This will be done by the following commits.
Liran Alon [Sat, 28 Mar 2020 20:00:56 +0000 (23:00 +0300)]
OvmfPkg/PvScsiDxe: Setup requests and completions rings
These rings are shared memory buffers between host and device in which
a cyclic buffer is managed to send request descriptors from host to
device and receive completion descriptors from device to host.
Note that because device may be constrained by IOMMU or guest may be run
under AMD SEV, we make sure to map these rings to device by using
PciIo->Map().
Liran Alon [Sat, 28 Mar 2020 20:00:49 +0000 (23:00 +0300)]
OvmfPkg/PvScsiDxe: Report the number of targets and LUNs
Implement EXT_SCSI_PASS_THRU.GetNextTarget() and
EXT_SCSI_PASS_THRU.GetNextTargetLun().
ScsiBusDxe scans all MaxTarget * MaxLun possible devices.
This can take unnecessarily long for large number of targets.
To deal with this, VirtioScsiDxe has defined PCDs to limit the
MaxTarget & MaxLun to desired values which gives sufficient
performance. It is very important in virtio-scsi as it can have
very big MaxTarget & MaxLun.
Even though a common PVSCSI device has a default MaxTarget=64 and
MaxLun=0, we implement similar mechanism as virtio-scsi for completeness.
This may be useful in the future when PVSCSI will have bigger values
for MaxTarget and MaxLun.
Liran Alon [Sat, 28 Mar 2020 20:00:47 +0000 (23:00 +0300)]
OvmfPkg/PvScsiDxe: Probe PCI devices and look for PvScsi
PvScsiControllerSupported() is called on handles passed in
by the ConnectController() boot service and if the handle is the
PVSCSI controller, the function would return success. A success
return value will attach our driver to the device.
Problem has been identified with Ip6ProcessRouterAdvertise() when
Router Advertise packet contains options with malicious/invalid
'Length' field. This can lead to platform entering infinite loop
when processing options from that packet.
Cc: Jiaxin Wu <jiaxin.wu@intel.com> Cc: Siyuan Fu <siyuan.fu@intel.com> Signed-off-by: Maciej Rabeda <maciej.rabeda@linux.intel.com> Reviewed-by: Siyuan Fu <siyuan.fu@intel.com>
Building OVMF for X64 with secure boot enabled on VS2019 results in
the following error:
d:\a\1\s\OvmfPkg\Library\GenericQemuLoadImageLib\GenericQemuLoadImageLib.c(154):
error C2220: the following warning is treated as an error
d:\a\1\s\OvmfPkg\Library\GenericQemuLoadImageLib\GenericQemuLoadImageLib.c(154):
warning C4244: '=': conversion from 'UINTN' to 'UINT32', possible loss of data
On enabling the /analyse option the VS2017 compiler
reports: warning C6001: Using uninitialized memory.
This warning is reported as some variables that were
being logged were uninitialised. To fix this, moved
the logging code after the variables being logged are
initialised.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com>
On enabling the /analyse option the VS2017 compiler
reports: warning C6001: Using uninitialized memory.
This warning is reported as some variables that were
being logged were uninitialised. To fix this, moved
the logging code after the variables being logged are
initialised.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com>
Sami Mujawar [Tue, 9 Jul 2019 15:50:42 +0000 (16:50 +0100)]
DynamicTablesPkg: Fix IORT node length assignment
The VS2017 compiler reports 'warning C4267: 'return': conversion
from 'size_t' to 'UINT32', possible loss of data' for a number of
functions that compute the IORT node length. Similarly, it reports
warnings for IORT node length field assignments as the length
field is 16-bit wide.
This patch adds type casts at appropriate places and also implements
validations to ensure that the max width of the respective fields
is not exceeded.
This patch also fixes a typo in one of the local variable names.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com>
Sami Mujawar [Tue, 9 Jul 2019 15:18:27 +0000 (16:18 +0100)]
DynamicTablesPkg: Serial debug port initialisation
The ARM DCC serial port subtype is an option that is
supported by the DBG2 generator. However, the serial
port initialisation should only be done for PL011/SBSA
compatible UARTs.
Add check to conditionally initialise the serial port.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com> Reviewed-by: Philippe Mathieu-Daude <philmd@redhat.com>
Sami Mujawar [Tue, 9 Jul 2019 14:58:37 +0000 (15:58 +0100)]
DynamicTablesPkg: Fix unaligned pointers usage
The VS2017 compiler reports 'warning C4366: The result of
the unary '&' operator may be unaligned' if an address of
an unaligned structure member is passed as an argument to
a function.
Fix this warning by using local variables in place of
unaligned structure members.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com>
Sami Mujawar [Tue, 9 Jul 2019 14:11:49 +0000 (15:11 +0100)]
DynamicTablesPkg: Fix ACPI table rev field width
The VS2017 compiler reports 'warning C4244: '=': conversion from
'const UINT32' to 'UINT8', possible loss of data' when the ACPI
table revision field is being updated.
The width of the revision field in the EFI_ACPI_DESCRIPTION_HEADER
struct is 8-bit wide. Therefore, to fix the above warning make the
ACPI Table revision field usage 8-bit wide across Dynamic Tables
Framework.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com>
Sami Mujawar [Tue, 9 Jul 2019 14:04:45 +0000 (15:04 +0100)]
DynamicTablesPkg: Fix Boot arch flag width
The ArmBootArch field of the FADT table is 16-bit wide. The
VS2017 compiler reports 'warning C4244: '=': conversion from
'UINT32' to 'UINT16', possible loss of data' when assigning the
CM_ARM_BOOT_ARCH_INFO.BootArchFlags value as the width of this
field in CM_ARM_BOOT_ARCH_INFO is 32-bit wide.
To fix this warning, update the CM_ARM_BOOT_ARCH_INFO struct
to make the BootArchFlags field 16-bit wide. This also makes
it compatible with the ACPI FADT specification.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com>
Sami Mujawar [Tue, 9 Jul 2019 12:54:04 +0000 (13:54 +0100)]
DynamicTablesPkg: Fix GT Block length assignment
The VS2017 compiler reports 'warning C4267: '=': conversion from
'size_t' to 'UINT16', possible loss of data'.
The sizeof() operator is used to calculate the size of the
GT Block structure. The length field in the GT Block structure
is 16-bit wide. Since the return type of sizeof() operator
is size_t the VS2017 compiler reports the above warning.
To fix the warning, an explicit type cast is added. An additional
check is also performed to ensure that the calculated GT Block
length does not exceed MAX_UINT16.
Signed-off-by: Sami Mujawar <sami.mujawar@arm.com> Reviewed-by: Alexei Fedorov <Alexei.Fedorov@arm.com> Reviewed-by: Philippe Mathieu-Daude <philmd@redhat.com>
projects / people / aperard / ovmf.git / log