edk2.git
3 days agoArmPkg/ArmGicDxe ARM: fix encoding for GICv3 interrupt acknowledge master
Ard Biesheuvel [Wed, 14 Nov 2018 19:27:24 +0000 (11:27 -0800)] 
ArmPkg/ArmGicDxe ARM: fix encoding for GICv3 interrupt acknowledge

Fix a typo in the 32-bit ARM version of the GICv3 driver, which uses
the wrong system register encoding to access ICC_IAR1, and attempted
to access ICC_IAR0 instead. This results in boot time hangs both
under QEMU emulation and on real hardware.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
3 days agoMdePkg/BaseIoLibIntrinsicArmVirt ARM: avoid double word loads and stores
Ard Biesheuvel [Wed, 7 Nov 2018 13:13:01 +0000 (14:13 +0100)] 
MdePkg/BaseIoLibIntrinsicArmVirt ARM: avoid double word loads and stores

BaseIoLibIntrinsicArmVirt was created to prevent LTO from merging
accesses to MMIO regions, resulting in instructions with multiple
output registers that KVM on ARM cannot emulate (since the exception
syndrome information that KVM relies on can only describe a single
output register)

However, using double word loads on ARM amounts to the same thing,
and so code that relies on doing 64-bit MMIO to regions that are
emulated under KVM (such as the GICv3 TYPER register) will still
suffer from the original issue.

So replace ldrd and strd with equivalent two instruction sequences.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
4 days agoUefiCpuPkg/CommonFeature: Always set FEATURE_CONTROL.Lock edk2-stable201811
Ruiyu Ni [Tue, 13 Nov 2018 07:34:33 +0000 (15:34 +0800)] 
UefiCpuPkg/CommonFeature: Always set FEATURE_CONTROL.Lock

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1305

The patch reverts commit 1ed6498c4a0210204bf4b95cc0c0cd6623ad6a0b
* UefiCpuPkg/CommonFeature: Skip locking when the feature is disabled

FEATURE_CONTROL.Lock bit is controlled by feature
CPU_FEATURE_LOCK_FEATURE_CONTROL_REGISTER. The commit 1ed649 fixes
a bug that when the feature is disabled, the Lock bit is cleared.
But it's a security hole if the bit is cleared when booting OS.
We can argue that platform needs to make sure the value
of PcdCpuFeaturesUserConfiguration should be set properly to make
sure feature CPU_FEATURE_LOCK_FEATURE_CONTROL_REGISTER is enabled.

But it's better to guarantee this in the generic core code.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Cc: Andrew Fish <afish@apple.com>
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
5 days agoMdeModulePkg/NvmExpressPei: Refine data buffer & len check in PassThru
Hao Wu [Fri, 9 Nov 2018 07:14:08 +0000 (15:14 +0800)] 
MdeModulePkg/NvmExpressPei: Refine data buffer & len check in PassThru

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=1142

The fix is similar to commit ebb6c7633bca47fcd5b460a67e18e4a717ea91cc.
We found that a similar fix should be applied to the NVMe PEI driver as
well. Hence, this one is for the PEI counterpart driver.

According to the the NVM Express spec Revision 1.1, for some commands
(like Get/Set Feature Command, Figure 89 & 90 of the spec), the Memory
Buffer maybe optional although the command opcode indicates there is a
data transfer between host & controller (Get/Set Feature Command, Figure
38 of the spec).

Hence, this commit refine the checks for the 'TransferLength' and
'TransferBuffer' field of the
EDKII_PEI_NVM_EXPRESS_PASS_THRU_COMMAND_PACKET structure to address this
issue.

Cc: Andrew Fish <afish@apple.com>
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Jiewen Yao <Jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Philippe Mathieu-Daude <philmd@redhat.com>
6 days agoMaintainers.txt: Update EDK II Releases to EDK-II-Release-Planning wiki
Liming Gao [Sat, 10 Nov 2018 14:05:58 +0000 (22:05 +0800)] 
Maintainers.txt: Update EDK II Releases to EDK-II-Release-Planning wiki

In V2, add Liming Gao as the maintainer of EDK II Releases.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
6 days agoReapply "OvmfPkg/QemuVideoDxe: list "UnalignedIoInternal.h" in the INF file"
Laszlo Ersek [Fri, 9 Nov 2018 19:16:21 +0000 (20:16 +0100)] 
Reapply "OvmfPkg/QemuVideoDxe: list "UnalignedIoInternal.h" in the INF file"

This reverts commit e038bde2679bbd200086c25ab43090ad3b8b25a3, reapplying
b2959e9f1a57279506ca46d56bc424fd7fa6b62a.

Note that the commit now being reverted is technically correct; the only
reason we're reverting it is because it should not have been pushed past
the Soft Feature Freeze for the edk2-stable201811 tag.

Cc: Anthony Perard <anthony.perard@citrix.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Julien Grall <julien.grall@linaro.org>
Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Cc: yuchenlin <yuchenlin@synology.com>
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1319
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: yuchenlin <yuchenlin@synology.com>
6 days agoReapply "OvmfPkg/QemuVideoDxe: VMWare SVGA device support"
Laszlo Ersek [Fri, 9 Nov 2018 19:15:31 +0000 (20:15 +0100)] 
Reapply "OvmfPkg/QemuVideoDxe: VMWare SVGA device support"

This reverts commit 98856a724c2acdc0094220d4de615a557dad0f88, reapplying
c137d95081690d4877fbeb5f1856972e84ac32f2.

Note that the commit now being reverted is technically correct; the only
reason we're reverting it is because it should not have been pushed past
the Soft Feature Freeze for the edk2-stable201811 tag.

Cc: Anthony Perard <anthony.perard@citrix.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Julien Grall <julien.grall@linaro.org>
Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Cc: yuchenlin <yuchenlin@synology.com>
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1319
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: yuchenlin <yuchenlin@synology.com>
6 days agoReapply "OvmfPkg/QemuVideoDxe: Helper functions for unaligned port I/O."
Laszlo Ersek [Fri, 9 Nov 2018 19:13:54 +0000 (20:13 +0100)] 
Reapply "OvmfPkg/QemuVideoDxe: Helper functions for unaligned port I/O."

This reverts commit 438ada5aa5a1174940795678c2dae07cde8f3869,
reapplying 05a5379458725234de8a05780fcb5da2c12680e4.

Note that the commit now being reverted is technically correct; the only
reason we're reverting it is because it should not have been pushed past
the Soft Feature Freeze for the edk2-stable201811 tag.

Cc: Anthony Perard <anthony.perard@citrix.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Julien Grall <julien.grall@linaro.org>
Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Cc: yuchenlin <yuchenlin@synology.com>
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1319
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: yuchenlin <yuchenlin@synology.com>
6 days agoReapply "OvmfPkg: VMWare SVGA display device register definitions"
Laszlo Ersek [Fri, 9 Nov 2018 18:43:55 +0000 (19:43 +0100)] 
Reapply "OvmfPkg: VMWare SVGA display device register definitions"

This reverts commit 328409ce8de7f318ee9c929b64302bd361cd1dbd, reapplying
9bcca53fe466cdff397578328d9d87d257aba493.

Note that the commit now being reverted is technically correct; the only
reason we're reverting it is because it should not have been pushed past
the Soft Feature Freeze for the edk2-stable201811 tag.

Cc: Anthony Perard <anthony.perard@citrix.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Julien Grall <julien.grall@linaro.org>
Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Cc: yuchenlin <yuchenlin@synology.com>
Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1319
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: yuchenlin <yuchenlin@synology.com>
7 days agoIntelFrameworkModulePkg: Fix UEFI and Tiano Decompression logic issue
Liming Gao [Thu, 8 Nov 2018 23:58:16 +0000 (07:58 +0800)] 
IntelFrameworkModulePkg: Fix UEFI and Tiano Decompression logic issue

https://bugzilla.tianocore.org/show_bug.cgi?id=1317

This is a regression issue caused by 684db6da64bc7b5faee4e1174e801c245f563b5c.
In Decode() function, once mOutBuf is fully filled, Decode() should return.
Current logic misses the checker of mOutBuf after while() loop.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
7 days agoMdePkg BaseUefiDecompressLib: Fix UEFI Decompression logic issue
Liming Gao [Thu, 8 Nov 2018 23:58:15 +0000 (07:58 +0800)] 
MdePkg BaseUefiDecompressLib: Fix UEFI Decompression logic issue

https://bugzilla.tianocore.org/show_bug.cgi?id=1317

This is a regression issue caused by 2ec7953d49677142c5f7552e9e3d96fb406ba0c4.
In Decode() function, once mOutBuf is fully filled, Decode() should return.
Current logic misses the checker of mOutBuf after while() loop.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
7 days agoBaseTools: Fix UEFI and Tiano Decompression logic issue
Liming Gao [Thu, 8 Nov 2018 23:58:14 +0000 (07:58 +0800)] 
BaseTools: Fix UEFI and Tiano Decompression logic issue

https://bugzilla.tianocore.org/show_bug.cgi?id=1317

This is a regression issue caused by 041d89bc0f0119df37a5fce1d0f16495ff905089.
In Decode() function, once mOutBuf is fully filled, Decode() should return.
Current logic misses the checker of mOutBuf after while() loop.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
7 days agoRevert "MdeModulePkg/DisplayEngine: Remove useless NULL ptr check for NewPos"
Liming Gao [Sat, 10 Nov 2018 14:06:39 +0000 (22:06 +0800)] 
Revert "MdeModulePkg/DisplayEngine: Remove useless NULL ptr check for NewPos"

This reverts commit 8cd4e734ccdfbc961c72aeaa8dbd3f5154171f9b.

It is not a real bug fix. It should not be pushed after
Hard Feature Freeze for edk2-stable201811 tag.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Bi Dandan <dandan.bi@intel.com>
7 days agoMdePkg: Fix incorrect check for DisplayOnly text format in AcpiEx
Dandan Bi [Thu, 8 Nov 2018 13:33:53 +0000 (21:33 +0800)] 
MdePkg: Fix incorrect check for DisplayOnly text format in AcpiEx

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1312

Text format for AcpiEx device path in UEFI Spec:
AcpiEx(HID,CID,UID,HIDSTR,CIDSTR,UIDSTR)
AcpiEx(HID|HIDSTR,(CID|CIDSTR,UID|UIDSTR))(Display Only)

When convert device path to text for AcpiEx device path,
current code check AllowShortcuts parameter to convert
the device path to DisplayOnly text format(shorter text
representation) by mistake.
It should check DisplayOnly parameter.

This commit is to fix this issue.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
7 days agoUefiCpuPkg/PiSmmCpuDxeSmm: Separate semaphore container.
Eric Dong [Sat, 10 Nov 2018 02:53:41 +0000 (10:53 +0800)] 
UefiCpuPkg/PiSmmCpuDxeSmm: Separate semaphore container.

In current implementation, core and package level sync uses same semaphores.
Sharing the semaphore may cause wrong execution order.
For example:
1. Feature A has CPU_FEATURE_CORE_BEFORE dependency with Feature B.
2. Feature C has CPU_FEATURE_PACKAGE_AFTER dependency with Feature B.
The expected feature initialization order is A B C:
A ---- (Core Depends) ----> B ---- (Package Depends) ----> C

For a CPU has 1 package, 2 cores and 4 threads. The feature initialization
order may like below:

   Thread#1             Thread#2       Thread#3         Thread#4
   [A.Init]             [A.Init]                        [A.Init]
Release(S1, S2)        Release(S1, S2)                Release(S3, S4)
Wait(S1) * 2           Wait(S2) * 2  <------------------------------- Core sync

   [B.Init]             [B.Init]
Release (S1,S2,S3,S4)
Wait (S1) * 4  <----------------------------------------------------- Package sync
                                                      Wait(S4 * 2) <- Core sync
                                                        [B.Init]

In above case, for thread#4, when it syncs in core level, Wait(S4) * 2 isn't
blocked and [B.Init] runs. But [A.Init] hasn't run in thread#3. It's wrong!
Thread#4 should execute [B.Init] after thread#3 executes [A.Init] because B
core level depends on A.

The reason of the wrong execution order is that S4 is released in thread#1
by calling Release (S1, S2, S3, S4) and in thread #4 by calling
Release (S3, S4).

To fix this issue, core level sync and package level sync should use separate
semaphores.

In above example, the S4 released in Release (S1, S2, S3, S4) should not be the
same semaphore as that in Release (S3, S4).

Related BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=1311

Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
7 days agoUefiCpuPkg/RegisterCpuFeaturesLib: Separate semaphore container.
Eric Dong [Thu, 8 Nov 2018 01:06:33 +0000 (09:06 +0800)] 
UefiCpuPkg/RegisterCpuFeaturesLib: Separate semaphore container.

In current implementation, core and package level sync uses same semaphores.
Sharing the semaphore may cause wrong execution order.
For example:
1. Feature A has CPU_FEATURE_CORE_BEFORE dependency with Feature B.
2. Feature C has CPU_FEATURE_PACKAGE_AFTER dependency with Feature B.
The expected feature initialization order is A B C:
A ---- (Core Depends) ----> B ---- (Package Depends) ----> C

For a CPU has 1 package, 2 cores and 4 threads. The feature initialization
order may like below:

   Thread#1             Thread#2       Thread#3         Thread#4
   [A.Init]             [A.Init]                        [A.Init]
Release(S1, S2)        Release(S1, S2)                Release(S3, S4)
Wait(S1) * 2           Wait(S2) * 2  <------------------------------- Core sync

   [B.Init]             [B.Init]
Release (S1,S2,S3,S4)
Wait (S1) * 4  <----------------------------------------------------- Package sync
                                                      Wait(S4 * 2) <- Core sync
                                                        [B.Init]

In above case, for thread#4, when it syncs in core level, Wait(S4) * 2 isn't
blocked and [B.Init] runs. But [A.Init] hasn't run in thread#3. It's wrong!
Thread#4 should execute [B.Init] after thread#3 executes [A.Init] because B
core level depends on A.

The reason of the wrong execution order is that S4 is released in thread#1
by calling Release (S1, S2, S3, S4) and in thread #4 by calling
Release (S3, S4).

To fix this issue, core level sync and package level sync should use separate
semaphores.

In above example, the S4 released in Release (S1, S2, S3, S4) should not be the
same semaphore as that in Release (S3, S4).

Related BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=1311

Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
7 days agoUefiCpuPkg/RegisterCpuFeaturesLib: Adjust Order.
Eric Dong [Fri, 9 Nov 2018 05:20:41 +0000 (13:20 +0800)] 
UefiCpuPkg/RegisterCpuFeaturesLib: Adjust Order.

V2 changes:
V1 change has regression which caused by change feature order.
V2 changes logic to detect dependence not only for the
neighborhood features. It need to check all features in the list.

V1 Changes:
In current code logic, only adjust feature position if current
CPU feature position not follow the request order. Just like
Feature A need to be executed before feature B, but current
feature A registers after feature B. So code will adjust the
position for feature A, move it to just before feature B. If
the position already met the requirement, code will not adjust
the position.

This logic has issue when met all below cases:
1. feature A has core or package level dependence with feature B.
2. feature A is register before feature B.
3. Also exist other features exist between feature A and B.

Root cause is driver ignores the dependence for this case, so
threads may execute not follow the dependence order.

Fix this issue by change code logic to adjust feature position
for CPU features which has dependence relationship.

Related BZ: https://bugzilla.tianocore.org/show_bug.cgi?id=1311

Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <Ruiyu.ni@intel.com>
10 days agoMdeModulePkg/DisplayEngine: Remove useless NULL ptr check for NewPos
Dandan Bi [Wed, 7 Nov 2018 08:14:22 +0000 (16:14 +0800)] 
MdeModulePkg/DisplayEngine: Remove useless NULL ptr check for NewPos

In function UiDisplayMenu, the NewPos ptr which used to point to the
highlight menu entry. It will always point to the menu entry which
need to be highlighted or the gMenuOption menu if the highlight menu
is not found.
So we can remove the NULL ptr check for NewPos in this function.
And add the ASSERT code to avoid if any false positive reports
of NULL pointer dereference issue raised from static analysis.

Cc: Liming Gao <liming.gao@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
10 days agoSourceLevelDebugPkg/Readme.txt: Update UDK Debugger Tool & manual URL
Hao Wu [Wed, 7 Nov 2018 02:27:28 +0000 (10:27 +0800)] 
SourceLevelDebugPkg/Readme.txt: Update UDK Debugger Tool & manual URL

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=1289

The commit updates the link to get access to the Intel(R) UDK Debugger
Tool and its detailed user manual.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
10 days agoShellPkg/ShellLib: Fix potential NULL deference issue
Ruiyu Ni [Wed, 7 Nov 2018 09:34:20 +0000 (17:34 +0800)] 
ShellPkg/ShellLib: Fix potential NULL deference issue

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1310

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Jim Dailey <jim_dailey@dell.com>
Reviewed-by: Hao A Wu <hao.a.wu@intel.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
10 days agoMdePkg/BaseSynchronizationLib: Fix InternalSync[De|In]crement
Ruiyu Ni [Wed, 7 Nov 2018 15:51:51 +0000 (23:51 +0800)] 
MdePkg/BaseSynchronizationLib: Fix InternalSync[De|In]crement

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1303

Today's code generates assembly code as below for
InternalSyncIncrement:
  __asm__ __volatile__ (
    "movl    $1, %%eax  \n\t"
    "lock               \n\t"
    "xadd    %%eax, %1  \n\t"
    "inc     %%eax      \n\t"
    : "=a" (Result),          // %0
      "+m" (*Value)           // %1
    :                         // no inputs that aren't also outputs
    : "memory",
      "cc"
    );

 0:       55      pushl   %ebp
 1:       89 e5   movl    %esp, %ebp
 3:       8b 45 08        movl    8(%ebp), %eax
 6:       b8 01 00 00 00  movl    $1, %eax
 b:       f0      lock
 c:       0f c1 00        xaddl   %eax, _InternalSyncIncrement(%eax)
 f:       40      incl    %eax
10:       5d      popl    %ebp
11:       c3      retl

Line #3 and Line #6 both use EAX as destination register.
Line #c uses EAX and (EAX).

The output operand "=a" tells GCC that EAX is used for output.
But GCC only assumes that EAX will be used in the very last
instruction.

Per GCC document,
"Use the '&' constraint modifier on all output operands that must
 not  overlap an input. Otherwise, GCC may allocate the output
 operand in the same register as an unrelated input operand, on
 the assumption that the assembler code consumes its inputs before
 producing outputs. This assumption may be false if the assembler
 code actually consists of more than one instruction."

"=&a" should be used to tell GCC not use EAX before the assembly.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Michael Kinney <michael.d.kinney@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Philippe Mathieu-Daudé <philmd@redhat.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Tested-by: Laszlo Ersek <lersek@redhat.com>
Suggested-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Fixes: 8a94eb9283fa09a30f5f06f0c12cf0ee4e14fbcf
Fixes: 17634d026f968c404b039a8d8431b6389dd396ea

11 days agoUefiCpuPkg/SmmCpu: Block access-out only when static paging is used
Jiewen Yao [Sat, 3 Nov 2018 11:59:01 +0000 (19:59 +0800)] 
UefiCpuPkg/SmmCpu: Block access-out only when static paging is used

When static paging is disabled, page table for below 4GB is created
and page table for above 4GB is created dynamically in page fault
handler.

Today's implementation only allow SMM access-out to below types of
memory address no matter static paging is enabled or not:
1. Reserved, run time and ACPI NVS type
2. MMIO

But certain platform feature like RAS may need to access other types
of memory from SMM. Today's code blocks these platforms.
This patch simplifies the policy to only block when static paging
is used so that the static paging can be disabled in these platforms
to meet their SMM access-out need.

Setting PcdCpuSmmStaticPageTable to FALSE can disable the static
paging.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
11 days agoArmPkg/ArmGicDxe: Fix GICv3 interrupt routing mode bug
Ming Huang [Mon, 29 Oct 2018 04:57:08 +0000 (12:57 +0800)] 
ArmPkg/ArmGicDxe: Fix GICv3 interrupt routing mode bug

Setting GICD_IROUTERn.IRM and GICD_IROUTERn.{Aff3,Aff2,Aff1,Aff0}
at the same time is nonsensical (see 8.9.13 in the GICv3 spec, which
says of GICD_IROUTERn.IRM that "When this bit is set to 1,
GICD_IROUTER<n>.{Aff3,Aff2,Aff1,Aff0} are UNKNOWN"). There is also no
guarantee that IRM is implemented (see GICD_TYPER.No1N which indicates
whether the implementation supports this or not).

Let's thus not set this bit, as we want all SPIs to be delivered to the
same CPU, and not be broadcast to all of them.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ming Huang <ming.huang@linaro.org>
Acked-by: Marc Zyngier <marc.zyngier@arm.com>
11 days agoMdeModulePkg/Core: fix ineffective guard page issue
Jian J Wang [Wed, 7 Nov 2018 00:48:09 +0000 (08:48 +0800)] 
MdeModulePkg/Core: fix ineffective guard page issue

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1295

This issue originates from following patch which allows to enable
paging if PcdImageProtectionPolicy and PcdDxeNxMemoryProtectionPolicy
(in addition to PcdSetNxForStack) are set to enable related features.

  5267926134d17e86672b84fd57b438f05ffa68e1

Due to above change, PcdImageProtectionPolicy will be set to 0 by
default in many platforms, which, in turn, cause following code in
MdeModulePkg\Core\Dxe\Misc\MemoryProtection.c fail the creation of
notify event of CpuArchProtocol.

1138:  if (mImageProtectionPolicy != 0 ||
           PcdGet64 (PcdDxeNxMemoryProtectionPolicy) != 0) {
1139:  Status = CoreCreateEvent (
...
1142:             MemoryProtectionCpuArchProtocolNotify,
...
1145:             );

Then following call flow won't be done and Guard pages will not be
set as not-present in SetAllGuardPages() eventually.

   MemoryProtectionCpuArchProtocolNotify()
=> HeapGuardCpuArchProtocolNotify()
=> SetAllGuardPages()

The solution is removing the if(...) statement so that the notify
event will always be created and registered. This won't cause
unnecessary code execution because, in the notify event handler,
the related PCDs like

    PcdImageProtectionPolicy and
    PcdDxeNxMemoryProtectionPolicy

will be checked again before doing related jobs.

Cc: Star Zeng <star.zeng@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
11 days agoMdeModulePkg/Core: fill logic hole in MemoryProtectionCpuArchProtocolNotify
Jian J Wang [Sat, 3 Nov 2018 05:58:28 +0000 (13:58 +0800)] 
MdeModulePkg/Core: fill logic hole in MemoryProtectionCpuArchProtocolNotify

At the end of of MemoryProtectionCpuArchProtocolNotify there's cleanup
code to free resource. But at line 978, 994, 1005 the function returns
directly. This patch use "goto" to replace "return" to make sure the
resource is freed before exit.

1029:  CoreCloseEvent (Event);
1030:  return;

There's another memory leak after calling gBS->LocateHandleBuffer() in
the same function:

  Status = gBS->LocateHandleBuffer (
                  ByProtocol,
                  &gEfiLoadedImageProtocolGuid,
                  NULL,
                  &NoHandles,
                  &HandleBuffer
                  );

HandleBuffer is allocated in above call but never freed. This patch
will also add code to free it.

Cc: Star Zeng <star.zeng@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
11 days agoRevert "XhciDxe: Use common buffer for AsyncInterruptTransfer"
Star Zeng [Tue, 6 Nov 2018 13:47:39 +0000 (21:47 +0800)] 
Revert "XhciDxe: Use common buffer for AsyncInterruptTransfer"

There is concern at the thread
https://lists.01.org/pipermail/edk2-devel/2018-November/031951.html.
And the time point is a little sensitive as it is near edk2-stable201811.

This reverts commit 777920997152a2e68f664241f6080b64ff21edd6.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
11 days agoRevert "EhciDxe: Use common buffer for AsyncInterruptTransfer"
Star Zeng [Tue, 6 Nov 2018 13:45:25 +0000 (21:45 +0800)] 
Revert "EhciDxe: Use common buffer for AsyncInterruptTransfer"

There is concern at the thread
https://lists.01.org/pipermail/edk2-devel/2018-November/031951.html.
And the time point is a little sensitive as it is near edk2-stable201811.

This reverts commit 0cd645250306b244a5d6e0e293ed1786ec101641.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
11 days agoUefiCpuPkg/CommonFeature: Skip locking when the feature is disabled
Ruiyu Ni [Tue, 6 Nov 2018 09:04:39 +0000 (17:04 +0800)] 
UefiCpuPkg/CommonFeature: Skip locking when the feature is disabled

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1305

Today's code unconditionally sets the IA32_FEATURE_CONTROL.Lock to 1
no matter the feature is enabled or not.

The patch fixes this issue by only setting the Lock bit to 1 when
the feature is enabled.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
12 days agoRevert "OvmfPkg: VMWare SVGA display device register definitions"
yuchenlin [Fri, 2 Nov 2018 03:24:01 +0000 (11:24 +0800)] 
Revert "OvmfPkg: VMWare SVGA display device register definitions"

This reverts commit 9bcca53fe466cdff397578328d9d87d257aba493.

We reverted VMWare SVGA driver. We don't need these definitions too.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: yuchenlin <yuchenlin@synology.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
12 days agoRevert "OvmfPkg/QemuVideoDxe: Helper functions for unaligned port I/O."
yuchenlin [Fri, 2 Nov 2018 03:24:00 +0000 (11:24 +0800)] 
Revert "OvmfPkg/QemuVideoDxe: Helper functions for unaligned port I/O."

This reverts commit 05a5379458725234de8a05780fcb5da2c12680e4.

The VMWare SVGA display device implemented by Qemu (-vga vmware) uses
an I/O-type BAR which is laid out such that some register offsets are
not aligned to the read/write width with which they are expected to be
accessed. However, we reverted the initialization of VMWare SVGA device,
we don't need such unaligned I/O.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: yuchenlin <yuchenlin@synology.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
12 days agoRevert "OvmfPkg/QemuVideoDxe: VMWare SVGA device support"
yuchenlin [Fri, 2 Nov 2018 03:23:59 +0000 (11:23 +0800)] 
Revert "OvmfPkg/QemuVideoDxe: VMWare SVGA device support"

This reverts commit c137d95081690d4877fbeb5f1856972e84ac32f2.

The VMWare SVGA model now -- since commit 104bd1dc70 in QEMU --
falls back to stdvga (that is, Bochs) if we don't setup VMWare SVGA
FIFO.

To simplify QemuVideoDxe, we don't intend to implement the VMWare SVGA
FIFO setup feature. It means our current VMW SVGA driver code is
basically dead. To simplify the problem, we will replace the old
VMWare SVGA driver to Bochs interface. It should work on all QEMU
version.

The first step for using Bochs interface is to revert old driver.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: yuchenlin <yuchenlin@synology.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
12 days agoRevert "OvmfPkg/QemuVideoDxe: list "UnalignedIoInternal.h" in the INF file"
yuchenlin [Fri, 2 Nov 2018 03:23:58 +0000 (11:23 +0800)] 
Revert "OvmfPkg/QemuVideoDxe: list "UnalignedIoInternal.h" in the INF file"

This reverts commit b2959e9f1a57279506ca46d56bc424fd7fa6b62a.

The VMWare SVGA display device implemented by Qemu (-vga vmware) uses
an I/O-type BAR which is laid out such that some register offsets are
not aligned to the read/write width with which they are expected to be
accessed. However, we will revert the initialization of VMWare SVGA
device later, we don't need such unaligned I/O.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: yuchenlin <yuchenlin@synology.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Regression-tested-by: Laszlo Ersek <lersek@redhat.com>
12 days agoShellPkg/TftpDynamicCommand: Clarify the retry count option in command.
Jiaxin Wu [Mon, 5 Nov 2018 06:52:30 +0000 (14:52 +0800)] 
ShellPkg/TftpDynamicCommand: Clarify the retry count option in command.

[-c <retry count>] is to define the number of times to transmit request
packets and wait for a response. The default value is 6. But it doesn't
specify the behavior of zero value. Here, The patch is to clear that:
Set to zero also means to use the default value.

Cc: Carsey Jaben <jaben.carsey@intel.com>
Cc: Ye Ting <ting.ye@intel.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
12 days agoMdeModulePkg/Mtftp4Dxe: Fix invalid configuration of MTFTP local port.
Jiaxin Wu [Mon, 5 Nov 2018 06:46:57 +0000 (14:46 +0800)] 
MdeModulePkg/Mtftp4Dxe: Fix invalid configuration of MTFTP local port.

This patch is to fix the invalid setting of MTFTP local port. The
issue can be reproduced by tftp shell command by using [-l port]
option.

Cc: Ye Ting <ting.ye@intel.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
12 days agoOvmfPkg: Replace obsoleted network drivers from platform DSC/FDF.
Fu Siyuan [Tue, 30 Oct 2018 06:49:42 +0000 (14:49 +0800)] 
OvmfPkg: Replace obsoleted network drivers from platform DSC/FDF.

V2:
Add missed library instance for NetworkPkg iSCSI driver.

This patch replaces the MdeModulePkg TCP, PXE and iSCSI driver with those
ones in NetworkPkg. These 3 drivers in MdeModulePkg are not being actively
maintained and will be removed from edk2 master soon.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Anthony Perard <anthony.perard@citrix.com>
Cc: Julien Grall <julien.grall@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
12 days agoVlv2TbltDevicePkg: Replace obsoleted drivers from platform DSC/FDF.
Fu Siyuan [Tue, 30 Oct 2018 06:51:16 +0000 (14:51 +0800)] 
Vlv2TbltDevicePkg: Replace obsoleted drivers from platform DSC/FDF.

V2:
Add missed library instance for NetworkPkg iSCSI driver.

This patch replaces the MdeModulePkg TCP, PXE and iSCSI driver with those
ones in NetworkPkg. These 3 drivers in MdeModulePkg are not being actively
maintained and will be removed from edk2 master soon.

Cc: David Wei <david.wei@intel.com>
Cc: Mang Guo <mang.guo@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-by: David Wei <david.wei@intel.com>
13 days agoMaintainer.txt: Add Chao to be co-maintainer of SignedCapsulePkg
Zhang, Chao B [Mon, 5 Nov 2018 14:19:24 +0000 (22:19 +0800)] 
Maintainer.txt: Add Chao to be co-maintainer of SignedCapsulePkg

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Zhang, Chao B <chao.b.zhang@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
13 days agoArmPkg/OpteeLib: Fix compilation issues for ARM (32-bit)
Sumit Garg [Fri, 2 Nov 2018 06:53:14 +0000 (12:23 +0530)] 
ArmPkg/OpteeLib: Fix compilation issues for ARM (32-bit)

Correct usage of EFI_PHYSICAL_ADDRESS (always 64-bit) vs. pointers
(depending on architecture).

Reported-by: Leif Lindholm <leif.lindholm@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Sumit Garg <sumit.garg@linaro.org>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
13 days agoEmulatorPkg: Replace obsoleted network drivers from platform DSC/FDF.
Fu Siyuan [Tue, 30 Oct 2018 06:47:33 +0000 (14:47 +0800)] 
EmulatorPkg: Replace obsoleted network drivers from platform DSC/FDF.

This patch replaces the MdeModulePkg TCP, PXE and iSCSI driver with those
ones in NetworkPkg. These 3 drivers in MdeModulePkg are not being actively
maintained and will be removed from edk2 master soon.

Cc: Jordan Justen <jordan.l.justen@intel.com>
Cc: Andrew Fish <afish@apple.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
13 days agoNt32Pkg: Replace obsoleted network drivers from NT32 platform DSC/FDF.
Fu Siyuan [Tue, 30 Oct 2018 06:43:37 +0000 (14:43 +0800)] 
Nt32Pkg: Replace obsoleted network drivers from NT32 platform DSC/FDF.

This patch replaces the MdeModulePkg TCP, PXE and iSCSI driver with those
ones in NetworkPkg. These 3 drivers in MdeModulePkg are not being actively
maintained and will be removed from edk2 master soon.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Fu Siyuan <siyuan.fu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
13 days agoCorebootPayloadPkg: Remove EdkShellBinPkg in FDF
shenglei [Mon, 5 Nov 2018 03:08:07 +0000 (11:08 +0800)] 
CorebootPayloadPkg: Remove EdkShellBinPkg in FDF

Remove EdkShellBinPkg in CorebootPayloadPkg.fdf.
https://bugzilla.tianocore.org/show_bug.cgi?id=1108

v3:Remove FULL_BIN and change SHELL_TYPE from FULL_BIN
to UEFI_BIN.

Cc: Maurice Ma <maurice.ma@intel.com>
Cc: Prince Agyeman <prince.agyeman@intel.com>
Cc: Benjamin You <benjamin.you@intel.com>
Change-Id: I4db7068a3a1f68a1f6303079b73dc548c9feb2e3
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Shenglei Zhang <shenglei.zhang@intel.com>
Reviewed-by: Benjamin You <benjamin.you@intel.com>
13 days agoBaseTools: Use common cc flag for building PcdValueInit.
BobCF [Fri, 2 Nov 2018 05:56:38 +0000 (13:56 +0800)] 
BaseTools: Use common cc flag for building PcdValueInit.

V3:

Update the patch to avoid "reduce" function fail.

V2:
Support to extract the common cc flag from a
combined cc flag string. For example
MSFT:*_*_IA32_CC_FLAGS = /D DISABLE_NEW_DEPRECATED_INTERFACES /DDEF_IA32
MSFT:*_*_X64_CC_FLAGS  = /DDEF_X64 /D DISABLE_NEW_DEPRECATED_INTERFACES

V1:
Use common cc flags for building PcdValueInit. The common
cc flags include the cc flag which is under common arch and
under all build arches.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Bob Feng <bob.c.feng@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
13 days agoBaseTools build_rule.template: Update aslc rule for XCODE tool chain
Liming Gao [Fri, 2 Nov 2018 15:17:07 +0000 (23:17 +0800)] 
BaseTools build_rule.template: Update aslc rule for XCODE tool chain

Update aslc rule to rename the temp output file from .efi to .pecoff.
This change can avoid the conflict .efi file name in output directory.
One is the driver image, another is aslc temp output file.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
13 days agoShellPkg-ShellApp: Provide fully-qualified path to shell scripts
Jim.Dailey@dell.com [Mon, 29 Oct 2018 21:14:51 +0000 (05:14 +0800)] 
ShellPkg-ShellApp: Provide fully-qualified path to shell scripts

Provide fully-qualified path to shell scripts

Section 3.6.2 of version 2.2 of the shell specification requires that
the first positional argument (i.e. arg 0) of a shell script resolves
to "the full path name of the script itself."

Ensure that the startup script and any scripts launched by the shell
meet this requirement.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jim Dailey <jim_dailey@dell.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
13 days agoShellPkg-UefiShellLib: Add a function to fully-qualify paths
Jim.Dailey@dell.com [Mon, 29 Oct 2018 21:14:36 +0000 (05:14 +0800)] 
ShellPkg-UefiShellLib: Add a function to fully-qualify paths

Add a function to return a clean, fully-qualified version of some path.

This function handles a (possibly "dirty") input path that may or may
not include a file system reference.

If it does not include a file system reference, then if the input path
does not begin with a forward or backward slash, then the input path is
relative to the current working directory of the current file system.
Otherwise, it is an absolute path within the current file system.

If it does include a file system reference, it may be a reference to the
current or some other file system.  If the file system reference is not
immediately followed by a forward or backward slash, then the input path
is relative to the current working directory of the given file system.
Otherwise, it is an absolute path within the given file system.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jim Dailey <jim_dailey@dell.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
13 days agoUefiCpuPkg/MpInitLib: Rollback old change 2a5997f8.
Dong, Eric [Fri, 2 Nov 2018 08:27:48 +0000 (16:27 +0800)] 
UefiCpuPkg/MpInitLib: Rollback old change 2a5997f8.

In some special cases, after BSP sends Init-sipi-sipi signal
AP needs more time to start the Ap procedure. In this case
BSP may think AP has finished its task but in fact AP hasn't began
yet.

Rollback former change to keep the status which only be used
when AP really finished task.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
13 days agoMaintainers.txt: Change package maintainer of UefiCpuPkg
Ruiyu Ni [Fri, 2 Nov 2018 09:21:37 +0000 (17:21 +0800)] 
Maintainers.txt: Change package maintainer of UefiCpuPkg

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
13 days agoMaintainers.txt: Change package maintainer of IntelSiliconPkg
Ruiyu Ni [Fri, 2 Nov 2018 09:19:03 +0000 (17:19 +0800)] 
Maintainers.txt: Change package maintainer of IntelSiliconPkg

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
2 weeks agoBaseTools: Check the max size for string PCD.
Zhao, ZhiqiangX [Wed, 31 Oct 2018 10:35:47 +0000 (18:35 +0800)] 
BaseTools: Check the max size for string PCD.

According to PCD_DATABASE_INIT in
edk2\MdeModulePkg\Include\Guid\PcdDataBaseSignatureGuid.h,
the max size for string PCD should not exceed USHRT_MAX 65535(0xffff).

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: ZhiqiangX Zhao <zhiqiangx.zhao@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Cc: Bob Feng <bob.c.feng@intel.com>
Reviewed-by: Bob Feng <bob.c.feng@intel.com>
2 weeks agoMdePkg/Base.h: Implement BASE_CR() via OFFSET_OF().
Marvin.Haeuser@outlook.com [Wed, 31 Oct 2018 20:08:35 +0000 (04:08 +0800)] 
MdePkg/Base.h: Implement BASE_CR() via OFFSET_OF().

Replace the current NULL pointer dereference to retrieve Field's
offset with a call to OFFSET_OF().  This is implemented via
__builtin_offsetof for GCC and Clang, which eliminates UB caught by
Clang UndefinedBehaviorSanitizer.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Marvin Haeuser <Marvin.Haeuser@outlook.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
2 weeks agoBaseTools tools_def.template: Add GCC link script option in ASLDLINK_FLAGS
Liming Gao [Thu, 1 Nov 2018 14:35:29 +0000 (22:35 +0800)] 
BaseTools tools_def.template: Add GCC link script option in ASLDLINK_FLAGS

GCC link script is used to discard the unused section data from ELF image.
ASLDLINK_FLAGS requires it to remove the unnecessary section data, then
GenFw can be used to retrieve the correct data section from ELF image.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
2 weeks agoBaseTools: Add checking to EFI variable attribute.
Feng, Bob C [Thu, 1 Nov 2018 15:07:01 +0000 (23:07 +0800)] 
BaseTools: Add checking to EFI variable attribute.

This patch is going to add a check that only if the
attribute of a EFI variable include 'NV', it will be
added into PcdNvStoreDefaultValueBuffer.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Bob Feng <bob.c.feng@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
2 weeks agoMdeModulePkg: Remove the trailing space in three source files.
Liming Gao [Thu, 1 Nov 2018 13:36:09 +0000 (21:36 +0800)] 
MdeModulePkg: Remove the trailing space in three source files.

The recent changes in these three source files introduce the trailing space.
This patch removes them to follow edk2 coding style.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
2 weeks agoBaseTools/Eot: Remove a duplication code in EotMain class
Hess Chen [Wed, 31 Oct 2018 05:53:27 +0000 (13:53 +0800)] 
BaseTools/Eot: Remove a duplication code in EotMain class

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hess Chen <hesheng.chen@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
2 weeks agoBaseTools: Add special handle for '\' use in Pcd Value
zhijufan [Wed, 17 Oct 2018 05:57:01 +0000 (13:57 +0800)] 
BaseTools: Add special handle for '\' use in Pcd Value

V2: Follow PEP8 to not multiples import on one line

Case:
gEfiOzmosisPkgTokenSpaceGuid.PcdBootLogFolderPath|L"\\Logs\\"|VOID*|12

Fixes: https://bugzilla.tianocore.org/show_bug.cgi?id=1287
Cc: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Zhiju.Fan <zhijux.fan@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
2 weeks agoBaseTools: Fix merge in FDF parser for OptionROM override
Tomas Pilar [Mon, 29 Oct 2018 16:52:53 +0000 (00:52 +0800)] 
BaseTools: Fix merge in FDF parser for OptionROM override

Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Tomas Pilar <tpilar@solarflare.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
2 weeks agoMdeModulePkg/PiSmmIpl: Do not reset SMRAM to UC when CPU driver runs
Ruiyu Ni [Mon, 22 Oct 2018 09:03:07 +0000 (17:03 +0800)] 
MdeModulePkg/PiSmmIpl: Do not reset SMRAM to UC when CPU driver runs

Today's PiSmmIpl implementation initially sets SMRAM to WB to speed
up the SMM core/modules loading before SMM CPU driver runs.
When SMM CPU driver runs, PiSmmIpl resets the SMRAM to UC. It's done
in SmmIplDxeDispatchEventNotify(). COMM_BUFFER_SMM_DISPATCH_RESTART
is returned from SMM core that SMM CPU driver is just dispatched.

Since now the SMRR is widely used to control the SMRAM cache setting.
It's not needed to reset the SMRAM to UC anymore.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jiewen Yao <jiewen.yao@intel.com>
Cc: Michael Kinney <michael.d.kinney@intel.com>
2 weeks agoNetworkPkg/TlsDxe: Fix failure to process multiple TLS records.
Jiaxin Wu [Tue, 30 Oct 2018 03:30:08 +0000 (11:30 +0800)] 
NetworkPkg/TlsDxe: Fix failure to process multiple TLS records.

Ref: https://bugzilla.tianocore.org/show_bug.cgi?id=1290.

Current implementation failed to parse the multiple TLS record
messages due to the incorrect pointer of TLS record header. This
patch is to resolve that problem.

Cc: Ye Ting <ting.ye@intel.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
2 weeks agoNetworkPkg/Mtftp6Dxe: Remove the trailing white spaces.
Jiaxin Wu [Wed, 31 Oct 2018 08:29:24 +0000 (16:29 +0800)] 
NetworkPkg/Mtftp6Dxe: Remove the trailing white spaces.

Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Fu Siyuan <siyuan.fu@intel.com>
2 weeks agoFmpDevicePkg DSC: Add NOOPT target to verify NOOPT target
Liming Gao [Tue, 30 Oct 2018 07:50:09 +0000 (15:50 +0800)] 
FmpDevicePkg DSC: Add NOOPT target to verify NOOPT target

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
2 weeks agoBaseTools ConvertFceToStructurePcd: Fix the array value with empty string
Liming Gao [Tue, 30 Oct 2018 14:59:45 +0000 (22:59 +0800)] 
BaseTools ConvertFceToStructurePcd: Fix the array value with empty string

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
2 weeks agoCryptoPkg/BaseCryptLib: Fix potential integer overflow issue.
Long Qin [Wed, 24 Oct 2018 13:16:42 +0000 (21:16 +0800)] 
CryptoPkg/BaseCryptLib: Fix potential integer overflow issue.

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1275

The LookupFreeMemRegion() in RuntimeMemAllocate.c is used to look-up
free memory region for runtime resource allocation, which was designed
to support runtime authenticated variable service.
The ReqPages in this function is the required pages to be allocated,
which depends on the malloc() call in internal OpenSSL routines. The
direct offset subtractions on ReqPages may bring possible integer
overflow issue.

This patch is to add the extra parameter checks to remove this possible
overflow risk.

Cc: Ye Ting <ting.ye@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Long Qin <qin.long@intel.com>
Reviewed-by: Ye Ting <ting.ye@intel.com>
2 weeks agoUefiCpuPkg/RegisterCpuFeaturesLib: Fix build failure.
Eric Dong [Tue, 30 Oct 2018 03:02:11 +0000 (11:02 +0800)] 
UefiCpuPkg/RegisterCpuFeaturesLib: Fix build failure.

Build UefiCpuPkg with below configuration:
Architecture(s)  = IA32
Build target     = NOOPT
Toolchain        = VS2015x86

Below error info shows up:
DxeRegisterCpuFeaturesLib.lib(CpuFeaturesInitialize.obj) :
error LNK2001: unresolved external symbol __allmul

Valid mDependTypeStr type only have 5 items, use UINT32 type cast
to fix this error.

Cc: Dandan Bi <dandan.bi@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
2 weeks agoMdeModulePkg/UdfDxe: Memory free/use after free in ResolveSymlink()
Hao Wu [Tue, 30 Oct 2018 01:17:04 +0000 (09:17 +0800)] 
MdeModulePkg/UdfDxe: Memory free/use after free in ResolveSymlink()

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=1279

For function ResolveSymlink(), the below codes:

    if (CompareMem ((VOID *)&PreviousFile, (VOID *)Parent,
                    sizeof (UDF_FILE_INFO)) != 0) {
      CleanupFileInformation (&PreviousFile);
    }

    CopyMem ((VOID *)&PreviousFile, (VOID *)File, sizeof (UDF_FILE_INFO));

If the contents in 'PreviousFile' and 'File' are the same, call to
"CleanupFileInformation (&PreviousFile);" will free the buffers in 'File'
as well. This will lead to potential memory double free/use after free
issues.

This commit will add additional check to address the above issue.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Paulo Alcantara <palcantara@suse.de>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2 weeks agoMdeModulePkg/UdfDxe: Content check for 'File' in ResolveSymlink()
Hao Wu [Tue, 30 Oct 2018 01:11:57 +0000 (09:11 +0800)] 
MdeModulePkg/UdfDxe: Content check for 'File' in ResolveSymlink()

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=1279

The content within 'File' is the output data for ResolveSymlink(). This
commit will add checks to ensure the content in 'File' is valid.
Otherwise, possible null pointer dereference issue will occur during the
subsequent usage of the data returned by ResolveSymlink().

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Paulo Alcantara <palcantara@suse.de>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2 weeks agoMdeModulePkg/UdfDxe: Check 'Component Type' within a Path Component
Hao Wu [Tue, 30 Oct 2018 01:06:48 +0000 (09:06 +0800)] 
MdeModulePkg/UdfDxe: Check 'Component Type' within a Path Component

REF:https://bugzilla.tianocore.org/show_bug.cgi?id=1279

According to the ECMA-167 standard (3rd Edition - June 1997), Section
14.16.1.1, valid values are 1 to 5. All other values will be treated as a
corrupted volume.

This commit will add such check within function ResolveSymlink().

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Paulo Alcantara <palcantara@suse.de>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2 weeks agoArmPkg/OpteeLib: Add APIs to communicate with OP-TEE
Sumit Garg [Mon, 22 Oct 2018 06:29:36 +0000 (11:59 +0530)] 
ArmPkg/OpteeLib: Add APIs to communicate with OP-TEE

Add following APIs to communicate with OP-TEE pseudo/early TAs:
1. OpteeInit
2. OpteeOpenSession
3. OpteeCloseSession
4. OpteeInvokeFunc

Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Sumit Garg <sumit.garg@linaro.org>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
2 weeks agoUefiCpuPkg/PiSmmCpuDxeSmm: Fix ASSERT for success.
Marvin H?user [Sun, 28 Oct 2018 08:51:23 +0000 (16:51 +0800)] 
UefiCpuPkg/PiSmmCpuDxeSmm: Fix ASSERT for success.

Index is initialized to MAX_UINT16 as default failure value, which
is what the ASSERT is supposed to test for.  The ASSERT condition
however can never return FALSE for INT16 != int, as due to
Integer Promotion[1], Index is converted to int, which can never
result in -1.

Furthermore, Index is used as a for loop index variable inbetween its
initialization and the ASSERT, so the value is unconditionally
overwritten too.

Fix the ASSERT check to compare Index to its upper boundary, which it
will be equal to if the loop was not broken out of on success.

[1] ISO/IEC 9899:2011, 6.5.9.4

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Marvin Haeuser <Marvin.Haeuser@outlook.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
2 weeks agoUefiCpuPkg/MpInitLib: Fix ASSERT for success.
Marvin H?user [Sun, 28 Oct 2018 08:51:22 +0000 (16:51 +0800)] 
UefiCpuPkg/MpInitLib: Fix ASSERT for success.

Index is initialized to MAX_UINT16 as default failure value, which
is what the ASSERT is supposed to test for.  The ASSERT condition
however can never return FALSE for INT16 != int, as due to
Integer Promotion[1], Index is converted to int, which can never
result in -1.

Furthermore, Index is used as a for loop index variable inbetween its
initialization and the ASSERT, so the value is unconditionally
overwritten too.

Fix the ASSERT check to compare Index to its upper boundary, which it
will be equal to if the loop was not broken out of on success.

[1] ISO/IEC 9899:2011, 6.5.9.4

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Marvin Haeuser <Marvin.Haeuser@outlook.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
2 weeks agoMdeModulePkg/Core: fix an issue of potential NULL pointer access
Jian J Wang [Mon, 29 Oct 2018 08:20:44 +0000 (16:20 +0800)] 
MdeModulePkg/Core: fix an issue of potential NULL pointer access

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1286

This issue is introduced by bb685071c2602cf786ea84c69bbebf2158194a38.

The *MemorySpaceMap assigned with NULL (line 1710) value might be
accessed (line 1726/1730) without any sanity check. Although it won't
happen in practice because of line 1722, we still need to add check
against NULL to make static code analyzer happy.

1710  *MemorySpaceMap       = NULL;
....  ...
1722  if (DescriptorCount == *NumberOfDescriptors) {
....  ...
1726    Descriptor = *MemorySpaceMap;
....  ...
1730        BuildMemoryDescriptor (Descriptor, Entry);

Tests:
  Pass build and boot to shell.

Cc: Hao Wu <hao.a.wu@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
2 weeks agoMdeModulePkg/UsbBusDxe: Add missing "return NULL" in UsbCreateDesc()
Ruiyu Ni [Mon, 29 Oct 2018 03:54:45 +0000 (11:54 +0800)] 
MdeModulePkg/UsbBusDxe: Add missing "return NULL" in UsbCreateDesc()

When (Len < Offset) is TRUE, indicating the data to visit is beyond
the boundary, the error message is printed but the function doesn't
return NULL.

It's a typo when modifying the commit 4c034bf62.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
2 weeks agoIntelFsp2Pkg: Fixed potentially NULL pointer accessing
Chasel, Chiu [Fri, 26 Oct 2018 07:12:33 +0000 (15:12 +0800)] 
IntelFsp2Pkg: Fixed potentially NULL pointer accessing

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1280

When copying IDT table in SecMain, the pointer might be
NULL so added the check to fix it.

Test: Verified on internal platform and boots successfully.

Cc: Jiewen Yao <Jiewen.yao@intel.com>
Cc: Desimone Nathaniel L <nathaniel.l.desimone@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Chasel Chiu <chasel.chiu@intel.com>
2 weeks agoMdeModulePkg/UsbBusPei: Reject descriptor whose length is bad
Ruiyu Ni [Thu, 25 Oct 2018 10:09:46 +0000 (18:09 +0800)] 
MdeModulePkg/UsbBusPei: Reject descriptor whose length is bad

Today's implementation doesn't check whether the length of
descriptor is valid before using it.

The patch fixes this issue by syncing the similar fix to UsbBusDxe.
70c3c2370a2aefe71cf0f6c1a1e063f7d74e1d79
*MdeModulePkg/UsbBus: Reject descriptor whose length is bad

Additionally the patch also rejects the data when length is
larger than sizeof (PeiUsbDevice->ConfigurationData).

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
2 weeks agoMdeModulePkg/UsbBusPei: Fix out-of-bound read access to descriptors
Ruiyu Ni [Thu, 25 Oct 2018 09:07:12 +0000 (17:07 +0800)] 
MdeModulePkg/UsbBusPei: Fix out-of-bound read access to descriptors

Today's implementation reads the Type/Length field in the USB
descriptors data without checking whether the offset to read is
beyond the data boundary.

The patch fixes this issue by syncing the fix in commit
4c034bf62cbc1f3c5f4b5df25de97f0f528132b2
*MdeModulePkg/UsbBus: Fix out-of-bound read access to descriptors

ParsedBytes in UsbBusPei.GetExpectedDescriptor() is different from
Consumed in UsbBusDxe.UsbCreateDesc().
ParsedBytes is the offset of found descriptor while Consumed is
offset of next descriptor of found one.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
2 weeks agoNetworkPkg/Mtftp6Dxe: Correct the total received and saved block number.
Jiaxin Wu [Thu, 25 Oct 2018 07:16:13 +0000 (15:16 +0800)] 
NetworkPkg/Mtftp6Dxe: Correct the total received and saved block number.

The block returned from Mtftp6RemoveBlockNum is not the total received and
saved block number if it works in passive (Slave) mode.

The issue was exposed by the EMS test.

Cc: Ye Ting <ting.ye@intel.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Ye Ting <ting.ye@intel.com>
2 weeks agoMdeModulePke/Mtftp4Dxe: Correct the total received and saved block number.
Jiaxin Wu [Thu, 25 Oct 2018 07:31:10 +0000 (15:31 +0800)] 
MdeModulePke/Mtftp4Dxe: Correct the total received and saved block number.

The block returned from Mtftp4RemoveBlockNum is not the total received and
saved block number if it works in passive (Slave) mode.

The issue was exposed by the EMS test.

Cc: Ye Ting <ting.ye@intel.com>
Cc: Fu Siyuan <siyuan.fu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Wu Jiaxin <jiaxin.wu@intel.com>
Reviewed-by: Ye Ting <ting.ye@intel.com>
3 weeks agoMdeModulePkg EhciDxe: Use common buffer for AsyncInterruptTransfer
Star Zeng [Thu, 25 Oct 2018 08:20:34 +0000 (16:20 +0800)] 
MdeModulePkg EhciDxe: Use common buffer for AsyncInterruptTransfer

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1274

In current code, EhcMonitorAsyncRequests (timer handler) will do
unmap and map operations for AsyncIntTransfers to "Flush data from
PCI controller specific address to mapped system memory address".
EhcMonitorAsyncRequests
  EhcFlushAsyncIntMap
    PciIo->Unmap
      IoMmu->SetAttribute
    PciIo->Map
      IoMmu->SetAttribute

This may impact the boot performance.

Since the data buffer for EhcMonitorAsyncRequests is internal
buffer, we can allocate common buffer by PciIo->AllocateBuffer
and map the buffer with EfiPciIoOperationBusMasterCommonBuffer,
then the unmap and map operations can be removed.

///
/// Provides both read and write access to system memory by
/// both the processor and a bus master. The buffer is coherent
/// from both the processor's and the bus master's point of view.
///
EfiPciIoOperationBusMasterCommonBuffer,

Test done:
USB KB works normally.
USB disk read/write works normally.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Cc: Jian J Wang <jian.j.wang@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
3 weeks agoMdeModulePkg XhciDxe: Use common buffer for AsyncInterruptTransfer
Star Zeng [Thu, 25 Oct 2018 08:20:27 +0000 (16:20 +0800)] 
MdeModulePkg XhciDxe: Use common buffer for AsyncInterruptTransfer

V3:
Call XhcFreeUrb after XhcCreateTransferTrb fails in XhcCreateTrb.

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1274

In current code, XhcMonitorAsyncRequests (timer handler) will do
unmap and map operations for AsyncIntTransfers to "Flush data from
PCI controller specific address to mapped system memory address".
XhcMonitorAsyncRequests
  XhcFlushAsyncIntMap
    PciIo->Unmap
      IoMmu->SetAttribute
    PciIo->Map
      IoMmu->SetAttribute

This may impact the boot performance.

Since the data buffer for XhcMonitorAsyncRequests is internal
buffer, we can allocate common buffer by PciIo->AllocateBuffer
and map the buffer with EfiPciIoOperationBusMasterCommonBuffer,
then the unmap and map operations can be removed.

///
/// Provides both read and write access to system memory by
/// both the processor and a bus master. The buffer is coherent
/// from both the processor's and the bus master's point of view.
///
EfiPciIoOperationBusMasterCommonBuffer,

Test done:
USB KB works normally.
USB disk read/write works normally.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Cc: Jian J Wang <jian.j.wang@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
3 weeks agoMdeModulePkg EhciDxe: Extract new EhciInsertAsyncIntTransfer function
Star Zeng [Wed, 24 Oct 2018 11:27:22 +0000 (19:27 +0800)] 
MdeModulePkg EhciDxe: Extract new EhciInsertAsyncIntTransfer function

V3:
Match function parameter name and description between
EhciSched.c and EhciSched.h.

V2:
Add the missing "gBS->FreePool (Data);".

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1274

Extract new EhciInsertAsyncIntTransfer function from
EhcAsyncInterruptTransfer.

It is code preparation for following patch,
no essential functional change.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Cc: Jian J Wang <jian.j.wang@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
3 weeks agoMdeModulePkg XhciDxe: Extract new XhciInsertAsyncIntTransfer function
Star Zeng [Mon, 22 Oct 2018 08:29:03 +0000 (16:29 +0800)] 
MdeModulePkg XhciDxe: Extract new XhciInsertAsyncIntTransfer function

V3:
Match function parameter name and description between
XhciSched.c and XhciSched.h.

V2:
Add the missing "FreePool (Data);".
Remove the unnecessary indentation change.

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1274

Extract new XhciInsertAsyncIntTransfer function from
XhcAsyncInterruptTransfer.

It is code preparation for following patch,
no essential functional change.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Cc: Jian J Wang <jian.j.wang@intel.com>
Cc: Jiewen Yao <jiewen.yao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
3 weeks agoUefiCpuPkg/RegisterCpuFeaturesLib: Fix GCC build failure.
Dong, Eric [Sat, 27 Oct 2018 09:09:21 +0000 (17:09 +0800)] 
UefiCpuPkg/RegisterCpuFeaturesLib: Fix GCC build failure.

Cc: Liming Gao <liming.gao@intel.com>
Cc: Dandan Bi <dandan.bi@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
3 weeks agoMdeModulePkg/Core: correct one coding style
Jian J Wang [Sat, 27 Oct 2018 02:12:13 +0000 (10:12 +0800)] 
MdeModulePkg/Core: correct one coding style

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1284

Non-Boolean comparisons should use a compare operator
(==, !=, >, < >=, <=)

Cc: Dandan Bi <dandan.bi@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Reviewed-by: Dandan Bi <dandan.bi@intel.com>
3 weeks agoBaseTools: Sync the DevicePath Function update from MdePkg
Feng, YunhuaX [Fri, 26 Oct 2018 08:37:14 +0000 (16:37 +0800)] 
BaseTools: Sync the DevicePath Function update from MdePkg

Update the AcpiExp, USBxxx, Sata, VENDOR_DEVICE_PATH.

Cc: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Yunhua Feng <yunhuax.feng@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
3 weeks agoBaseTools: Rollback Filter out unused structure pcds
Feng, Bob C [Fri, 26 Oct 2018 14:34:52 +0000 (22:34 +0800)] 
BaseTools: Rollback Filter out unused structure pcds

This reverts commit 51d17bb7b0da0d9c9e91c226f1982d7020f43795.
commit 51d17bb7b0da0d9c9e91c226f1982d7020f43795 adds new check
of Pcds in the platform unused library INF files.
It breaks the existing platform.
To avoid the impact, roll back this change first.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Bob C Feng <bob.c.feng@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
3 weeks agoMdeModulePkg RegularExpressionDxe: Remove unknown build option for XCODE
Liming Gao [Fri, 26 Oct 2018 02:33:24 +0000 (10:33 +0800)] 
MdeModulePkg RegularExpressionDxe: Remove unknown build option for XCODE

This patch makes RegularExpressionDxe pass XCODE5 build.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Liming Gao <liming.gao@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
3 weeks agoBaseTools: Allow multiple PciDeviceId in Fdf OptionROM override
Tomas Pilar [Wed, 24 Oct 2018 11:42:36 +0000 (19:42 +0800)] 
BaseTools: Allow multiple PciDeviceId in Fdf OptionROM override

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Tomas Pilar <tpilar@solarflare.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
3 weeks agoBaseTools: Fix BPDG tool print traceback info issue
Feng, YunhuaX [Thu, 25 Oct 2018 05:51:01 +0000 (13:51 +0800)] 
BaseTools: Fix BPDG tool print traceback info issue

Fix BPDG tool print traceback info issue
and remove abundant code

Cc: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Yunhua Feng <yunhuax.feng@intel.com>
Reviewed-by: Yonghong Zhu <yonghong.zhu@intel.com>
3 weeks agoMdeModulePkg/Core: fix an IA32 build failure
Jian J Wang [Fri, 26 Oct 2018 04:45:07 +0000 (12:45 +0800)] 
MdeModulePkg/Core: fix an IA32 build failure

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1277

The failure is caused by data type conversion between UINTN and UINT64,
which is checked in at 63ebde8ef6d4ff497d054ccc010904ecd4441198.

Cc: Star Zeng <star.zeng@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Jian J Wang <jian.j.wang@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
3 weeks agoMdeModulePkg/HiiDB: Make sure database update behaviors are atomic
Dandan Bi [Fri, 12 Oct 2018 10:13:19 +0000 (18:13 +0800)] 
MdeModulePkg/HiiDB: Make sure database update behaviors are atomic

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1235

When update contents in HiiDatabase, like:
1. Add/update/remove package list
2. Add/update string
3. Add/update image
We should make these operations atomic to prevent the
potential issue that the one update operation with
higher TPL may interrupt another.
This commit is to make the HiiDatabase update behaviors
atomic by adding EfiAcquireLock/EfiReleaseLock function.

Cc: Liming Gao <liming.gao@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
3 weeks agoMdeModulePkg/HiiDB: Reorganize codes of exporting HII settings
Dandan Bi [Thu, 11 Oct 2018 06:37:09 +0000 (14:37 +0800)] 
MdeModulePkg/HiiDB: Reorganize codes of exporting HII settings

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1235

Function "HiiGetConfigurationSetting" may be called
when HiiDatabase contents have been updated.
And it is just to call function "HiiGetDatabaseInfo" to
get the contents in HiiDatabase and call function
"HiiGetConfigRespInfo" and get the configuration response
string form HII drivers.

So here we can remove function "HiiGetConfigurationSetting"
and make caller to call "HiiGetDatabaseInfo" and
"HiiGetConfigRespInfo" directly.

And thus it also can distinguish which code blocks are to
operate HiiDatabase contents and which code blocks are not.
Then it's easy to know which code blocks should be atomic
when updating HiiDatabase contents.

Cc: Liming Gao <liming.gao@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
3 weeks agoBaseTools: Roll back code modify by commit 9e47e6f90880
Feng, YunhuaX [Fri, 26 Oct 2018 06:07:59 +0000 (14:07 +0800)] 
BaseTools: Roll back code modify by commit 9e47e6f90880

Roll back code modify by commit 9e47e6f90880,
if ForceRebase not False or True, the GenFv command not need add parameter

Cc: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Yunhua Feng <yunhuax.feng@intel.com>
3 weeks agoUefiCpuPkg/RegisterCpuFeaturesLib: Support combo CPU feature style.
Eric Dong [Thu, 25 Oct 2018 06:50:22 +0000 (14:50 +0800)] 
UefiCpuPkg/RegisterCpuFeaturesLib: Support combo CPU feature style.

Current code assume only one dependence (before or after) for one
feature. Enhance code logic to support feature has two dependence
(before and after) type.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
3 weeks agoIntelFsp2Pkg: Fix GCC49/XCODE build failure
Chasel, Chiu [Fri, 26 Oct 2018 02:41:49 +0000 (10:41 +0800)] 
IntelFsp2Pkg: Fix GCC49/XCODE build failure

REF: https://bugzilla.tianocore.org/show_bug.cgi?id=1276

Fixed potentially uninitialized variable build failure
caused by commit: b1cc6f672f3b924cdb190e5b92db3b47f46a8911

Test: Verified on internal platform and boots successfully.

Cc: Jiewen Yao <Jiewen.yao@intel.com>
Cc: Desimone Nathaniel L <nathaniel.l.desimone@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Chasel Chiu <chasel.chiu@intel.com>
Reviewed-by: Jiewen Yao <Jiewen.yao@intel.com>
3 weeks agoUefiCpuPkg/PiSmmCpuDxeSmm: Fix build failure for VS2012 and GCC49.
Eric Dong [Thu, 25 Oct 2018 02:12:52 +0000 (10:12 +0800)] 
UefiCpuPkg/PiSmmCpuDxeSmm: Fix build failure for VS2012 and GCC49.

Code initialized in function can't be correctly detected by build tool.
Add code to clearly initialize the local variable before use it.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Dandan Bi <dandan.bi@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
3 weeks agoUefiCpuPkg/PiSmmCpuDxeSmm: Clean up useless code.
Eric Dong [Thu, 25 Oct 2018 02:03:57 +0000 (10:03 +0800)] 
UefiCpuPkg/PiSmmCpuDxeSmm: Clean up useless code.

Remove useless code after change 93324390.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Dandan Bi <dandan.bi@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
3 weeks agoUefiCpuPkg/PiSmmCpuDxeSmm: Remove white space at line end.
Eric Dong [Thu, 25 Oct 2018 01:51:29 +0000 (09:51 +0800)] 
UefiCpuPkg/PiSmmCpuDxeSmm: Remove white space at line end.

Remove extra white space at the end of line.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Dandan Bi <dandan.bi@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
3 weeks agoUefiCpuPkg/RegisterCpuFeaturesLib: Fix ECC issues.
Eric Dong [Thu, 25 Oct 2018 01:51:00 +0000 (09:51 +0800)] 
UefiCpuPkg/RegisterCpuFeaturesLib: Fix ECC issues.

Changes include:
1. Remove extra white space at the end of line.
2. Add comments for the new add function parameter.
3. Update IN OUT tag for function parameter.

Cc: Dandan Bi <dandan.bi@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
3 weeks agoUefiCpuPkg/CpuCommonFeaturesLib: Remove white space at line end.
Eric Dong [Thu, 25 Oct 2018 01:49:35 +0000 (09:49 +0800)] 
UefiCpuPkg/CpuCommonFeaturesLib: Remove white space at line end.

Remove extra white space at the end of line.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Dandan Bi <dandan.bi@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
3 weeks agoUefiCpuPkg/RegisterCpuFeaturesLib: Fix build failure for VS2012 and GCC49.
Eric Dong [Thu, 25 Oct 2018 01:42:09 +0000 (09:42 +0800)] 
UefiCpuPkg/RegisterCpuFeaturesLib: Fix build failure for VS2012 and GCC49.

Code initialized in function can't be correctly detected by build tool.
Add code to clearly initialize the local variable before use it.

Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Cc: Dandan Bi <dandan.bi@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Eric Dong <eric.dong@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
3 weeks agoBaseTools:Not miss the full assign value of FixedAtBuild structure PCD
Zhaozh1x [Wed, 24 Oct 2018 12:31:23 +0000 (20:31 +0800)] 
BaseTools:Not miss the full assign value of FixedAtBuild structure PCD

For structure PCD, if it is a FixedAtBuild PCD, the full assign value in
dsc file should not be missed when updating the structure PCD value.

Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: ZhiqiangX Zhao <zhiqiangx.zhao@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Yonghong Zhu <yonghong.zhu@intel.com>
Cc: Bob Feng <bob.c.feng@intel.com>
Reviewed-by: Bob Feng <bob.c.feng@intel.com>