In edk2, the division and shifting of 64-bit values are forbidden with
C-language operators, because the compiler may generate intrinsic calls
for them.
For example, clang-3.8 emits a call to "__umoddi3" for
UDF_LOGICAL_SECTOR_SIZE % Media->BlockSize
in PartitionInstallUdfChildHandles(), if PartitionDxe is built for IA32,
which then fails to link.
UDF_LOGICAL_SECTOR_SIZE has type UINT64, while
EFI_BLOCK_IO_MEDIA.BlockSize has type UINT32(). Replace the % operator
with a DivU64x32Remainder() call.
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Paulo Alcantara <pcacjr@zytor.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
In edk2, initialization of local variables is forbidden, both for
stylistic reasons and because such initialization may generate calls to
compiler intrinsics.
For the following initialization in UdfRead():
CHAR16 FileName[UDF_FILENAME_LENGTH] = { 0 };
clang-3.8 generates a memset() call, when building UdfDxe for IA32, which
then fails to link.
Replace the initialization with ZeroMem().
Do the same to "FilePath" in UdfOpen().
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Paulo Alcantara <pcacjr@zytor.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Ard reports that clang-3.8 correctly flags the following issue in the
ReadFile() function:
If "RecordingFlags" is INLINE_DATA, then there are three paths through the
code where we mean to return success, but forget to set Status
accordingly:
(1) when "ReadFileInfo->Flags" is READ_FILE_GET_FILESIZE, or
(2) when "ReadFileInfo->Flags" is READ_FILE_ALLOCATE_AND_READ and
AllocatePool() succeeds, or
(3) when "ReadFileInfo->Flags" is READ_FILE_SEEK_AND_READ.
Set "Status" to EFI_SUCCESS when we are done processing the INLINE_DATA
request, i.e., when we reach the corresponding "break" statament under the
INLINE_DATA case label.
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Paulo Alcantara <pcacjr@zytor.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Reported-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Reviewed-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
In the ReadFile() function, if "RecordingFlags" is INLINE_DATA, then we
cover the following values of "ReadFileInfo->Flags":
- READ_FILE_GET_FILESIZE
- READ_FILE_ALLOCATE_AND_READ
- READ_FILE_SEEK_AND_READ
We don't do anything (just proceed to the end of the function) if
"ReadFileInfo->Flags" is anything else.
In reality the above three values cover the domain of the
UDF_READ_FILE_FLAGS enum type fully, and "ReadFileInfo->Flags" is only
ever set internally to UdfDxe. Therefore any other flag value would be a
bug in UdfDxe.
ASSERT() specifically that "ReadFileInfo->Flags" has been set correctly,
so that the reader is not left wondering what happens if none of the enum
constants match.
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Paulo Alcantara <pcacjr@zytor.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Initialize the array DescriptorLBAs[] after declaration to fix
non-constant aggregate initializer warning in VS tool chains.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Paulo Alcantara <pcacjr@zytor.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
Initialize the array DescriptorLBAs[] after declaration to fix
non-constant aggregate initializer warning in VS tool chains.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Paulo Alcantara <pcacjr@zytor.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
V3: Remove one unnecessay type cast in patch 1.
Codes:
if (FilePosition + ExtentLength > ReadFileInfo->FilePosition) {
Offset = ReadFileInfo->FilePosition - FilePosition;
if (Offset < 0) {
Offset = -(Offset)
}
...
}
Offset is UINT64 can not < 0, so the code logic may have some issue.
and Offset = -(Offset) may build failure in some circumstance.
Previously type cast Offset to INT64 to fix build break. Now remove
the type cast. Then can to check the code logic later.
Cc: Eric Dong <eric.dong@intel.com>
Cc: Paulo Alcantara <pcacjr@zytor.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Dandan Bi <dandan.bi@intel.com>
Reviewed-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
This patch introduces UDF file system support in EDK2. All block devices
that support BlockIo and DiskIo protocols and contain a valid UDF file
system - as specified by OSTA Universal Disk Format (revisions 1.02
through 2.60) - will be installed EFI_SIMPLE_FILE_SYSTEM_PROTOCOL to
provide access to underlying file system.
File system operations on regular, directory and symlink files are
supported.
Cc: Star Zeng <star.zeng@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Scan for UDF file systems on all block devices, as specified by OSTA
Universal Disk Format Specification 2.60, and install a Vendor-Defined
Media Device Path for each file system found.
The Vendor-Defined Media Device Path for the UDF file system is then
checked by UdfDxe to decide whether or not start the driver.
Cc: Star Zeng <star.zeng@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Laszlo Ersek <lersek@redhat.com>
Contributed-under: TianoCore Contribution Agreement 1.1
Signed-off-by: Paulo Alcantara <pcacjr@zytor.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
RamDiskDxe installs the RamDiskAcpiCheck() Ready To Boot callback
function. If EFI_ACPI_TABLE_PROTOCOL and/or EFI_ACPI_SDT_PROTOCOL are not
found, then informational messages are logged, and the RAM disks are not
published to the (nonexistent) NFIT table.
The logic is fine, but the info messages are not concatenated correctly
from multiple string literals -- the second parts are passed as (unused)
arguments to DEBUG(). Fix the typos.
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Hao Wu <hao.a.wu@intel.com>
Cc: Leif Lindholm <leif.lindholm@linaro.org>
Cc: Star Zeng <star.zeng@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Star Zeng <star.zeng@intel.com>
This PCD is used to indicated the recovery file name.
The previous name - FvMain.Fv is hardcoded in CdExpressPei.
It does not make sense to force the name.
Now a platform may use any recovery file name.
Cc: Feng Tian <feng.tian@intel.com>
Cc: Star Zeng <star.zeng@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Michael D Kinney <michael.d.kinney@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Michael D Kinney <michael.d.kinney@intel.com>
Current PeiCdExpress driver will return capsule size to
be block aligned.
It will fail if we check the capsuleImageSize field.
The patch correct the returned capsule size.
Tested-by: Eric Dong <eric.dong@intel.com>
Cc: Feng Tian <feng.tian@intel.com>
Cc: Eric Dong <eric.dong@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jiewen Yao <jiewen.yao@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
This commit will scan all the EFI raw section instances within the
module's FV to make sure the NVDIMM root device SSDT can be properly
located.
Cc: Feng Tian <feng.tian@intel.com>
Cc: Zhang Chao B <chao.b.zhang@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Chao Zhang <chao.b.zhang@intel.com>
Adding an option in HII menu so user can choose memory type to use when
creating a RAM Disk in system.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Tapan Shah <tapandshah@hpe.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Reviewed-by: Samer El-Haj-Mahmoud <elhaj@hpe.com>
The 'Size' numeric value used when creating a raw RAM disk does not
require a varstore to save its previous value in the create raw RAM disk
HII page.
The expecting behavior is that after a user created a raw RAM disk, the
next time when the create raw RAM disk page is entered, the 'Size' numeric
will be the default value (EFI_PAGE_SIZE).
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Reviewed-by: Tapan Shah <tapandshah@hpe.com>
Potentially uninitialized local variables 'TableKey' and 'TableHeader'
might be used in functions RamDiskPublishNfit() and RamDiskUnpublishNfit()
in file RamDiskProtocol.c.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Qiu Shumin <shumin.qiu@intel.com>
The RamDiskDxe driver originally uses a variable-length HII varstore to
retrieve the HII checkbox status of each registered RAM disk.
However, HII does not support the variable-length varstore feature.
Therefore, only the checkbox status for the first 8 RAM disks are tracked
for the following definition of HII varstore structure considering the
alignment:
typedef struct {
UINT64 Size;
UINT8 RamDiskList[0];
} RAM_DISK_CONFIGURATION;
This commit uses the private data of each registered RAM disks to track
the HII checkbox status instead to resolve the issue.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
The RamDiskDxe now will report RAM disks with reserved memory type to NFIT
in the ACPI table.
This commit will also make sure that an NVDIMM root device exists in the
\SB scope before reporting any RAM disk to NFIT.
To properly report the NVDIMM root device, one will need to append the
following content in the [Rule.Common.DXE_DRIVER] field in platform FDF
files:
RAW ACPI Optional |.acpi
RAW ASL Optional |.aml
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Reviewed-by: Samer El-Haj-Mahmoud <elhaj@hpe.com>
EndingAddress is calculated incorrectly. Original code calculates it as
(StartingAddress + Size). Correct value should be (StartingAddress +
Size - 1.
Note:
Besides the changes made by Samer, Hao also fixed a similar issue in
RamDiskImpl.c
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Samer El-Haj-Mahmoud <elhaj@hpe.com>
Signed-off-by: Tapan Shah <tapandshah@hpe.com>
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Tapan Shah <tapandshah@hpe.com>
There are scenario when the BindingStop service of PartitionDxe driver be
re-entered.
An example will be ejecting a DVD from a SATA DVDROM and then run
"reconnect -r" under shell. In this specific case, part of the calling
stack will be:
PartitionDriverBindingStop() (PartitionDxe) ->
Stop first child handle (PartitionDxe) ->
ScsiDiskFlushBlocksEx() (ScsiDiskDxe) ->
A media change is detected (ScsiDiskDxe) ->
Reinstall of BlockIO(2) protocols (ScsiDiskDxe) ->
Entering PartitionDriverBindingStop() again (PartitionDxe) ->
Potential risk of referencing already stopped child handle (PartitionDxe)
...
The current code has potential issue of referencing of already stopped
child handle. This commit adds re-entry handling logic to resolve such
issue.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Found an issue that file system cannot be started on a DVD when doing the
following process:
Boot to shell with a DVD inside a SATA DVDROM. Eject the DVD and run
"reconnect -r". Put the DVD inside again and run "reconnect -r".
The cause is that after executing the second reconnect action, DiskIo
immediately returns EFI_NO_MEDIA in function DiskIo2ReadWriteDisk() when
checking the media information. However, at this time, the media
information does not get updated by the ScsiDisk driver. Therefore, DiskIo
driver should left the no media check to ScsiDisk driver.
Generally, the media changed and media write protect check should also be
left to lower-level device driver. Thus, these two checks in function
DiskIo2ReadWriteDisk() are also removed.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Some ISO images cannot be recognized properly when they are on media whose
block size is not 2048 bytes.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
There is no asynchronous operations to the registered RAM disks link list
maintained within RamDiskDxe driver, therefore, the TPL raise and restore
operations when dealing with the link list are unnecessary.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Previously, the code uninstalls the DEVICE_PATH_PROTOCOL with the device
path given by caller of the 'RamDiskUnregister' function. The given device
path might be different from the one used to install the
DEVICE_PATH_PROTOCOL.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
The 'DisconnectController' function calls after
'UninstallMultipleProtocolInterfaces' are unnecessary, since
'DisconnectController' is called inside function
'UninstallMultipleProtocolInterfaces'.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
The RamDiskDxe driver will:
1. Produce the EFI RAM Disk Protocol
2. Install RAM disk device path and block I/O related protocols on the
RAM disk device handle.
3. Install RAM disk configuration form to HII database
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Samer El-Haj-Mahmoud <elhaj@hpe.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
UEFI Spec 2.3.1.D Section 9.3.5.1 says Partitions are numbered
according to their entry in their respective partition table,
starting with 1.
Update the code for MBR partition to follow this rule.
Ref: https://github.com/tianocore/edk2/issues/50
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
Add check to see if there is a media in the disk device. If not return
EFI_NO_MEDIA directly.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Hao Wu <hao.a.wu@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19446 6f19259b-4bc3-4df7-8a09-765794883524
The call in ProbeMediaStatus() to the ReadDisk() function of the
EFI_DISK_IO_PROTOCOL interface implemented in DiskIoDxe/DiskIo.c
crashed in DiskIo2ReadWriteDisk() because of the NULL value of
the destination buffer pointer.
Pass the address of a buffer in the stack instead of a NULL
pointer.
In addition to avoiding the crash, that way, the media probe does not
depend anymore on the way the EFI_DISK_IO_PROTOCOL implementation deals
with a NULL value of the destination buffer pointer as the UEFI
specification does not specify the expected behaviour.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ronald Cron <Ronald.Cron@arm.com>
Reviewed-by: Olivier Martin <olivier.martin@arm.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@17859 6f19259b-4bc3-4df7-8a09-765794883524
El Torito format can be used on different media (eg: USB).
A ISO image can be dumped onto a USB mass-storage.
These media might not have the same block size as the CDROM media (ie: 2KB).
The El Torito code and the specification assumes a LBA 2KB.
In addition, the specification says in "12.3.4.4 CD-ROM and DVD-ROM":
UEFI code does not assume a fixed block size.
I was able to dupliacte the issue by copying a debian ISO on a USB driver.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Olivier Martin <olivier.martin@arm.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@16600 6f19259b-4bc3-4df7-8a09-765794883524
DiskIoDxe: Fix ReadDiskEx and WriteDiskEx to not modify the user’s buffer when the BufferSize is 0.
DiskIoDxe: Fix ReadDiskEx and WriteDiskEx hang issue when the submitted blockio2 task is completed before submitting another blockio2 task.
DiskIoDxe: Fix FlushEx to free the flush task item in callback (memory leak issue).
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@16215 6f19259b-4bc3-4df7-8a09-765794883524
1. Module UNI and Package UNI files are not DOS format. Convert them to DOS format.
2. Remove unused SectionExtractionDxeModStrs.uni and SectionExtractionPeiModStrs.uni
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Gao, Liming <liming.gao@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@16044 6f19259b-4bc3-4df7-8a09-765794883524
2. Add MODULE_UNI_FILE file that contains the localized Abstract and Description of a module.
a. Addresses an information gap between INF files and the UEFI Distribution Packaging Specification XML schema
b. There will be an associated update to UPT in BaseTools to consume MODULE_UNI_FILE and associated UNI file during UDP creation that performs the INF -> XML conversion.
c. There will be an associated update to UPT in BaseTools to produce MODULE_UNI_FILE and associated UNI file during UDP installation that performs the XML -> INF conversion.
3. Add Module Extra UNI file that provides the localized Name of a module.
a. [UserExtensions.TianoCore."ExtraFiles"] provides an easy method for a module to specify extra files not listed in [Sources] or [Binaries] sections to be added to a UDP without having to list the files in the UPT package information data file.
b. There will be an associated update to UPT in BaseTools to package up files listed in [UserExtensions.TianoCore."ExtraFiles"] during UDP creation.
c. UNI file contains localized name of a module to go along with the localized Abstract and Description from the MODULE_UNI_FILE.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Zeng, Star <star.zeng@intel.com>
Reviewed-by: Gao, Liming <liming.gao@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@15963 6f19259b-4bc3-4df7-8a09-765794883524
1. Usage information in INF file comment blocks are either incomplete or incorrect.
This includes usage information for Protocols/PPIs/GUIDs/PCDs/HOBs/Events/BootModes.
The syntax for usage information in comment blocks is defined in the EDK II Module Information (INF) Specification
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Zeng, Star <star.zeng@intel.com>
Reviewed-by: Gao, Liming <liming.gao@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@15962 6f19259b-4bc3-4df7-8a09-765794883524
PcdDiskIoDataBufferBlockNum replaced the hardcoded value into
the Disk I/O driver.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Olivier Martin <olivier.martin@arm.com>
Reviewed-By: Tian, Feng <feng.tian@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@15235 6f19259b-4bc3-4df7-8a09-765794883524
1. DiskIo2 shouldn't signal the event when the *Ex interface returns failure status per the UEFI spec.
2. PartitionDxe should close DiskIo2 protocol when error happens in DriverBindingStart() otherwise Fat driver cannot open the DiskIo2 BY_DRIVER.
3. PartitionDxe should create event using TPL_NOTIFY instead of TPL_CALLBACK otherwise asynchronous FileIo may be blocked.
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Feng Tian <feng.tian@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@14680 6f19259b-4bc3-4df7-8a09-765794883524
DiskIoCreateSubtaskList() may split the transfer into three segments:
- a leading segment, called underrun, which is the fractional, trailing
subset of the first underlying block,
- a middle segment, which is an integral multiple of underlying blocks,
- a trailing segment, called overrun, which is the fractional, leading
subset of the last underlying block.
This is an example read from the /EFI/BOOT/BOOTX64.EFI file, on the
RHEL-6.4 installation ISO (debug log enabled with EFI_D_BLKIO). The
underlying block size is 2048 bytes (IDE CD-ROM).
DiskIo: Create subtasks for task: Offset/BufferSize/Buffer = 0000000000004600/00002000/BD890018
R:Lba/Offset/Length/WorkingBuffer/Buffer = 0000000000000008/00000600/00000200/BD90D000/BD890018
R:Lba/Offset/Length/WorkingBuffer/Buffer = 000000000000000C/00000000/00000600/BD90D000/BD890218
R:Lba/Offset/Length/WorkingBuffer/Buffer = 0000000000000009/00000000/00001800/00000000/BD890218
The first line corresponds to the underrun.
The second line corresponds to the overrun.
The third line corresponds to the middle segment.
In decimal:
- task: read 8192 bytes from offset 17920, storing it at BD890018
- underrun:
- read block 8 [16384..18432) into the transfer area,
- copy 512 bytes from offset 1536 of the transfer area to BD890018
(target buffer offset 0, running total: 512)
- middle segment:
- read blocks 9, 10, 11 [18432..24576) into the transfer area,
- copy 6144 bytes from offset 0 of the transfer area to BD890218
(target buffer offset 512, running total: 6656)
- overrun:
- read block 12 [24576..26624) into the transfer area,
- copy 1536 bytes from offset 0 of the transfer area to BD890218 (!!!)
(target buffer offset 512 (!!!), running total 8192)
The values marked with (!!!) constitute the bug --
DiskIoCreateSubtaskList() doesn't take the size of the middle segment into
account when it calculates the destination (for reads) or source (for
writes) pointer for the overrun. This leads to data corruption.
When reading, data is copied form the transfer area to the target buffer
with
CopyMem (Subtask->Buffer, Subtask->WorkingBuffer + Subtask->Offset, Subtask->Length);
calls in DiskIo2OnReadWriteComplete() for nonblocking reads, and in
DiskIo2ReadWriteDisk() for blocking reads. Therefore it's enough to adjust
Subtask->Buffer when it is initialized. (See BD891A18 below.)
DiskIo: Create subtasks for task: Offset/BufferSize/Buffer = 0000000000004600/00002000/BD890018
R:Lba/Offset/Length/WorkingBuffer/Buffer = 0000000000000008/00000600/00000200/BD90D000/BD890018
R:Lba/Offset/Length/WorkingBuffer/Buffer = 000000000000000C/00000000/00000600/BD90D000/BD891A18
R:Lba/Offset/Length/WorkingBuffer/Buffer = 0000000000000009/00000000/00001800/00000000/BD890218
The patched call to DiskIoCreateSubtask() is also executed for write
requests. The changed Subtask->Buffer initialization fixes the "overrun
half writes" in DiskIo2ReadWriteDisk() too:
//
// A sub task before this one should be a block read operation, causing
// the WorkingBuffer filled with the entire one block data.
//
CopyMem (Subtask->WorkingBuffer + Subtask->Offset, Subtask->Buffer, Subtask->Length);
This code doubles for underrun and overrun half-writes. The patch doesn't
modify the underrun case.
If we're storing the overrun at the beginning of the pre-read last block
(which we're going to write out as a full block), then
- Subtask->Offset == 0,
- Subtask->Length == OverRun,
- the first byte *not* accessed in the source area is
((Buffer + UnderRunLength) + BufferSize) + OverRun.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@14602 6f19259b-4bc3-4df7-8a09-765794883524
Error: Variable "Media" was set but never used.
Change-Id: I1297b3de797e436ed403375c801945b3202b532c
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Olivier Martin <olivier.martin@arm.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@14575 6f19259b-4bc3-4df7-8a09-765794883524
Change DiskIoDxe to produce DiskIo2 protocol when the BlockIo2 protocol is available.
Change PartitionDxe to produce BlockIo2 protocol based on DiskIo2 protocol instead of BlockIo2 protocol.
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Kinney Michael <michael.d.kinney@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@14570 6f19259b-4bc3-4df7-8a09-765794883524
Let the driver binding version of Partition driver be higher than that of Fat driver to make sure the DriverBindingStart() of the Partition driver gets run before that of Fat driver so that all the partitions can be recognized.
Signed-off-by: niruiyu
Reviewed-by: erictian
git-svn-id: https://edk2.svn.sourceforge.net/svnroot/edk2/trunk/edk2@12528 6f19259b-4bc3-4df7-8a09-765794883524
Fixed NULL pointer reference of BlockIo2.Media in partition code. This resulted
in the interrupt vector table being corrupted in a legacy (CSM) based system.
Signed-off-by: geekboy15a
Reviewed-by: mdkinney
Reviewed-by: niruiyu
Reviewed-by: jljusten
git-svn-id: https://edk2.svn.sourceforge.net/svnroot/edk2/trunk/edk2@11711 6f19259b-4bc3-4df7-8a09-765794883524
The others changes include:
a. Check RemainingDevicePath at beginning of Supported(), make sure it has been verified before Start() is called.
b. Check IO protocol firstly rather than EfiDevicePathProtocolGuid, reduce the times entering into Start() function because EfiDevicePathProtocolGuid existed on most of handle.
2. roll back serial drivers not to create child device, if the device speicifed by remainingdevicepath cannot find in the created devices list.
git-svn-id: https://edk2.svn.sourceforge.net/svnroot/edk2/trunk/edk2@9267 6f19259b-4bc3-4df7-8a09-765794883524
EFI_SIGNATURE_16 -> SIGNATURE_16
EFI_SIGNATURE_32 -> SIGNATURE_32
EFI_SIGNATURE_64 -> SIGNATURE_64
EFI_FIELD_OFFSET -> OFFSET_OF
EFI_MAX_BIT -> MAX_BIT
EFI_MAX_ADDRESS -> MAX_ADDRESS
These macros are not defined in UEFI spec. It makes more sense to use the equivalent macros in Base.h to avoid alias.
git-svn-id: https://edk2.svn.sourceforge.net/svnroot/edk2/trunk/edk2@7056 6f19259b-4bc3-4df7-8a09-765794883524
Laszlo Ersek
Dandan Bi
Paulo Alcantara
Hao Wu
Gary Lin
Jiewen Yao
Tapan Shah
Samer El-Haj-Mahmoud
Alcantara, Paulo
Ni, Ruiyu
Ruiyu Ni
Jordan Justen
Ronald Cron
Feng Tian
Olivier Martin
Samer El-Haj-Mahmoud
Gao, Liming
Zeng, Star
Sergey Isakov
jyao1
niruiyu
qianouyang
jljusten
lzeng14
geekboy15a
ydong10
erictian
hhtian
lgao4
qhuang8
vanjeff
eric_tian
gikidy
jji4
klu2
xli24