Commit 9168df3dea
"ShellPkg/ShellProtocol.c: Handle memory allocation failure"
only keeps the protocol clean up in CleanUpShellProtocol() and
creates a new function CleanUpShellEnvironment() which calls
CleanUpShellProtocol(), then unregisters the hotkey callback.
But the commit forgot to change the Shell.c to call
CleanUpShellEnvironment() which causes the hotkey callback is
not unregistered while the callback function doesn't exist
when Shell exits.
This causes system hang when pressing CTRL+C after exiting shell.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Eric Dong <eric.dong@intel.com>
And add Shell prefix to the two library APIs.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
1) RunSplitCommand() allocates the initial SplitStdOut via
CreateFileInterfaceMem(). Free SplitStdIn after the swap to fix
the memory leak.
2) In RunSplitCommand(), SplitStdOut is checked for equality with
StdIn. This cannot happen due to the if-check within the swap.
Hence remove it.
3) UefiMain() doesn't free SplitList. Delete all list entries and
reinitialize the list when in DEBUG. This does not include the
CreateFileInterfaceMem()-allocated SplitStd mentioned in 1), so
keep the ASSERT() until resolved.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Marvin Haeuser <Marvin.Haeuser@outlook.com>
Reviewed-by: Qiu Shumin <shumin.qiu@intel.com>
The EFI_UNICODE_COLLATION_PROTOCOL can have two different GUIDs.
Look for both to support more UEFI implementations.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Marvin Haeuser <Marvin.Haeuser@outlook.com>
Reviewed-by: Qiu Shumin <shumin.qiu@intel.com>
If 'ReadKeyStroke' function return EFI_NOT_READY then skip it.
If the return value is EFI_DEVICE_ERROR clean the currentString buffer.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Pedroa Liu <pedroa.liu@insyde.com>
Reviewed-by: Qiu Shumin <shumin.qiu@intel.com>
UEFI Shell 2.x cannot recognize whether a .EFI file is an application or
a driver. This means when we typed in a driver image in Shell command
line, Shell will load the driver image and try to run the entry point
function of the driver.
This patch check the ImageCodeType to fix the issue.
Cc: Jaben Carsey <jaben.carsey@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
the performance.
Currently UEFI Shell reads variable storage to get the environment
variables every time running a new command. And reading(writing)
UEFI variables is a high cost operation on most platforms. In order
to enhance the performance this patch read the variable storage once
and cache the environment variables in memory. Every further 'set'
command will save the variable not only to Shell cache, but also the
flash variable storage.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
Reviewed-by:Ruiyu Ni <ruiyu.ni@intel.com>
The BuildOptions for AARCH64 in Application/Shell/Shell.inf only affect
the core Shell binary, and not the Shell component libraries which are
merged into the final Shell binary via NULL library class resolution.
This means we need to override the UEFI_APPLICATION build options in
the platform .DSC anyway, there is no point in setting these options
here as well. So remove them.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Liming Gao <liming.gao@intel.com>
Reviewed-by: Leif Lindholm <leif.lindholm@linaro.org>
This patch makes Shell support -nonesting invocation option. This option
specifies that EFI_SHELL_PROTOCOL.Execute API nesting of a new Shell
instance is optional and dependent on the 'nonesting' Shell environment
variable.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jaben Carsey <jaben.carsey@intel.com>
Reviewed-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Jim Dailey <Jim_Dailey@Dell.com>
Use DOS format end of line(CR, LF).
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
If data of any real size were to be piped from one command to another,
an inordinate amount of time could be taken up by reallocating memory
that is only 10 bytes bigger than what is currently needed. Also, this
could cause unwelcome memory fragmentation.
Added a define to control how much memory is reallocated beyond that
which is currently needed. Set it to 1K vs. the original 10 bytes.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jim Dailey <jim_dailey@dell.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
The shell uses the memory file structure to manage temporary files in
memory that support piping of output from one command into the the
input of another command. The BufferSize member is the size of the
internal buffer, not the size of the data that was written to the
file. So, it was possible to read beyond the EOF of these files as
reads used BufferSize. Now FileSize tracks the actual size of these
files (the number of bytes written, not the number of bytes available
in the buffer), and the reads use this member.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jim Dailey <jim_dailey@dell.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
If the BOM is stripped from StdIn, then an app that duplicates StdIn
will not be able to duplicate, say, a UCS2 file that was piped into
it (the output file it creates would not start with a BOM).
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jim Dailey <jim_dailey@dell.com>
Reviewed-by: Jaben Carsey <Jaben.Carsey@intel.com>
UEFI Shell scandalizes the help message in spec level so that a standalone
UEFI shell application can never get "-?" switch, instead the Shell core
(interpreter) detects the "-?" and finds .MAN file for that shell
application in certain spec defined paths, then show the help extracted
from that .MAN file.
But it means distributing a UEFI shell application not only means
distributing a .EFI file but also distributing a .MAN file. If the text
formatted .MAN file is corrupted (edited by user by mistake), or is
missing (deleted by user by mistake), no help will be shown to user.
So this patch enhance the Shell to make it support finding help message
imbedded in resource section of application image.
Cc: Jaben Carsey <jaben.carsey@intel.com>
Cc: Ruiyu Ni <ruiyu.ni@intel.com>
Cc: Liming Gao <liming.gao@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Liming Gao <liming.gao@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
The pointer 'FileInterface->Buffer' returned from 'AllocateZeroPool' in function
'CreateFileInterfaceMem' may be NULL and will be dereferenced at the following code.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
Fix various errors when piping a UNICODE or ASCII file to a simple shell application that reads standard input and writes it to standard output.
1) When the memory file is created by CreateFileInferfaceMem() to capture the pipe output, no UNICODE BOM is written to the memory file. Later, when the memory file is read by the application using ShellFileHandleReadLine(), the function indicates that the file is ASCII because there is no BOM.
2) If the file is piped as ASCII, the ASCII memory image is not correctly created by FileInterfaceMemWrite() as each ASCII character is followed by '\0' in the image (when the ASCII data is written to the memory image, the file position should only be incremented by half the buffer size).
3) ShellFileHandleReadLine() does not read ASCII files correctly (writes to Buffer need to be cast as CHAR8*).
4) FileInterfaceMemRead() and FileInterfaceMemWrite() as somewhat hard to read and difficult to debug with certain tools due to the typecasting of This. Added a local variable (MemFile) of the correct type to these functions and used it instead of This.
Enhancement: ShellFileHandleReadLine() now returns EFI_END_OF_FILE when appropriate.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jim Dailey <jim_dailey@dell.com>
reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
When we run following script in Shell:
"
for %a run (1 200)
echo %a
memmap
endfor
"
We may find memory leak in system. This patch free buffer in 'BufferToFreeList' to avoid this issue.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19521 6f19259b-4bc3-4df7-8a09-765794883524
Add the PCD to PcdShellMaxHistoryCommandCount indicate the max count of history commands.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19520 6f19259b-4bc3-4df7-8a09-765794883524
This patch fixes the following compile error under GCC 5.3.1:
/home/pcacjr/work/edk2.git/ShellPkg/Application/Shell/ShellManParser.c:
In function 'IsTitleHeader':
/home/pcacjr/work/edk2.git/ShellPkg/Application/Shell/ShellManParser.c:502:5:
error: enumeration value 'Final' not handled in switch [-Werror=switch]
switch (State) {
^
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Paulo Alcantara <paulo.alc.cavalcanti@hp.com>
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19304 6f19259b-4bc3-4df7-8a09-765794883524
Prior to this change, the shell would not use a MAN file if the Title Header line was not strictly formatted.
For example, if the case of the command name in the file was not exactly the same as the case of the command name as typed by the user,
the MAN file would not be used. Also, extra whitespace on the line would also cause the shell to ignore the MAN file.
This change allows "extra" white space and ignores case when looking for the command name.
It also ignores any path information for cases where the user enters a relative or absolute path to the EFI file.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Jim Dailey <jim_dailey@dell.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19290 6f19259b-4bc3-4df7-8a09-765794883524
When the FvSimpleFileSystemDxe module is included in a platform,
Simple File System Protocols are produced for firmware volumes(FV)
that do not have the same style device paths as file systems with
file names. The ShellPkg has an assumption that the device path
contains device path nodes of type MEDIA_FILEPATH_DP and generates
an ASSERT() if any other device path nodes are encountered. This
change removes the ASSERT() condition and instead returns NULL that
means EfiShellGetFilePathFromDevicePath() can not convert the
device path nodes that represent the file path to a Unicode string.
Cc: Jaben Carsey <jaben.carsey@intel.com>
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Michael Kinney <michael.d.kinney@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19228 6f19259b-4bc3-4df7-8a09-765794883524
Building the latest shell on RVCT exposed this warning:
ShellPkg\Application\Shell\Shell.c(1090,69): error #191-D: type qualifier is meaningless on cast type
The CONST in the cast was deemed meaningless. Removing the CONST fixed the warning.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Cohen Eugene <eugene@hp.com>
Reviewed-by: Qiu Shumin <shumin.qiu@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18954 6f19259b-4bc3-4df7-8a09-765794883524
Per Shell SPEC 2.1 'Double-quotation marks that surround arguments are not stripped in positional parameters'. This patch makes Shell implementation to follow SPEC.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18742 6f19259b-4bc3-4df7-8a09-765794883524
1. Add a new function 'RunShellCommand' to return command status, thus 'EfiShellExecute' can get the command return status of 'CommandLine'.
2. Refine the code logic of 'EfiShellExecute' to make the new image of shell be loaded only if 'Environment' isn't NULL.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Signed-off-by: Jin Eric <eric.jin@intel.com>
Signed-off-by: Jaben Carsey <jaben.carsey@intel.com>
Reviewed-by: Ruiyu Ni <ruiyu.ni@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18664 6f19259b-4bc3-4df7-8a09-765794883524
In Shell spec 2.1 the return name of EFI_SHELL_PROTOCOL.GetCurDir() is defined as 'fs0:\current-dir' while in current implementation it's 'fs0:\current-dir\'.
To follow spec the patch removed the redundant '\' char.
Since it has been broken for a long time, some codes may depend on the broken behavior.
After this change 'EFI_SHELL_PROTOCOL.GetCurDir()' and 'UefiShellLib.ShellGetCurrentDir()'
will return a current directory string without tailing '\' (fs0:\current-dir), the value of Shell environment variable 'cwd' will become 'fs0:\current-dir' as well.
This patch has updated all the code in EDKII to make them depend on the new behavior.
Developers should check whether 'GetCurDir()' and 'ShellGetCurrentDir' are used in their source code.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18653 6f19259b-4bc3-4df7-8a09-765794883524
TestArgv.nsh is a very simple shell script to test how the interpreter parses
the parameters. It uses ShellCTestApp.efi to dump the parameters passed from the
interpreter.
TestArgv.log is the desired output created using "TestArgv.nsh > TestArgv.log".
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Ruiyu Ni <ruiyu.ni@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18558 6f19259b-4bc3-4df7-8a09-765794883524
1. Fix GCC build fail.
2. It's not correct to cast away constness to allow TrimSpaces() to modify 'commandline'.
This patch makes a copy of 'commandLine' and work with that in the remainder of the function.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18500 6f19259b-4bc3-4df7-8a09-765794883524
When execute a command with tailing blank spaces in ShellProtocol.Execute() Shell will fail.
This patch move the TrimSpaces operation into ParseCommandLineToArgs function to fix the problem.
Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Yang Jadis <jadis.yang@intel.com>
Signed-off-by: Qiu Shumin <shumin.qiu@intel.com>
Reviewed-by: Jaben Carsey <jaben.carsey@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@18491 6f19259b-4bc3-4df7-8a09-765794883524
Ruiyu Ni
Marvin H?user
Pedroa Liu
Jaben Carsey
Qiu Shumin
Liming Gao
Ard Biesheuvel
Jim_Dailey@Dell.com
Yao Jiewen
Paulo Alcantara
Michael Kinney
Jordan Justen
Cohen Eugene
Yang Jadis