The basic logical channel 0 is always present. It cannot be created or
closed. Let's restrict the value range of chan_nr, so that only valid
lchan numbers can be passed.
Related: OS#6531
Change-Id: I4eebd9f15fadd18e1caeb033fda36c59446fcab8
When we define command arguments using the ArgumentParser, we sometimes
define the positional arguments first. However, since positional arguments
usually follow after the optional (--xyz) arguments, we should define the
positional arguments last.
Related: OS#6531
Change-Id: I2412eb6e7dc32ae95a575f31d4489ce210d85ea0
The CardKeyProvider support for the commands enable_chv, disable_chv,
verify_chv, change_chv and unblock_chv is broken. The reason for this
is the annotation "type=is_decimal" in the argument parser. This annotation
prevents the usage of string placeholders ("PIN1", "PUK1", etc).
Let's fix this by finding a better solution. We can also replace any
missing PIN/PUK code by checking if it is supplied or not. If not,
we query the CardKeyProvider. This also makes the usage of the *_chv
commands more uniform with the verify_adm command.
Related: OS#6531
Change-Id: I565b56ac608e801c67ca53d337bdec9efa3f3817
In the previous patch we have improved the export command. Since
the implementation of the fsdump command is very similar to the
implementation of the export command we can now apply the same
improvements to the fsdump command as well.
Change-Id: I4d2ef7b383025a5bbf122f18ecd51b7d73aaba14
Related: OS#6092
Since we now have the ability to provide export methods for all file
types in the file system (this also includes DF and ADF files), we need
to support this at shell command level as well. Let's also renovate the
walk method and the action method that does the actual exporting.
Related: OS#6092
Change-Id: I3ee661dbae5c11fec23911775f352ac13bc2c6e5
the select_parent method in RuntimeLchan currently implements a way
to escape from an application that has no filesystem support. However,
this escape route can be integrated directly into the select_file
method. This will give us the benefit that it will work transparently
in all code locations.
(This also means we can get rid of the select_parent method again)
Related: OS#6120
Change-Id: Ie6f37d13af880d24a9c7a8a95cef436b603587c7
This command exports the entire filesystem state as one JSON document,
which can be useful for storing it in a noSQL database, or for doing a
structured diff between different such dumps.
It's similar to "export", but then reasonably different to rectify a
separate command.
Change-Id: Ib179f57bc04d394efe11003ba191dca6098192d3
The function do_switch_channel method calls methods in RuntimeLchan
that should be private. There is also a code duplication in
RuntimeLchan that should be cleaned up.
Related: OS#6092
Change-Id: Ie5e5f45787abaaf032e1b49f51d447653cf2c996
The reset command resets the card using the card object. This unfortunately
leaves the RuntimeState uninformed about the event. However, the RuntimeState
class also has a reset method that resets the card and the RuntimeState. Let's
use this reset method. Also fix this method so that it ensures that the SCP is
also no longer present.
Related: OS#6092
Change-Id: I1ad29c9e7ce7d80bebc92fa173ed7a44ee4c2998
When we traverse the file system using the command "export" we will
also select all ADFs but not all ADFs may have UICC file system support.
This makes it impossible to exit those ADFs again. To exit anyway we
select an application with filesystem support first and then the parent
EF we wanted to select originally. This method may not only be useful
when traversing the filesystem, so let's put it into the RuntimeLchan
class and change it a little so that it would also work if the ADF in
question is an a sub DF.
Related: OS#6092
Change-Id: I72de51bc7519fafbcc71d829719a8af35d774342
The code that generates the filesystem export lines for the various
different file structures can be moved into the filesystem class model.
This simplifies the code since we do not need any extra logic to
distinguish between the different file structures.
Related: OS#6092
Change-Id: Icc2ee60cfc4379411744ca1033d79a1ee9cff5a6
It's generally a bad idea to keep [card specific] key material lying
around unencrypted in CSV files. The industry standard solution in the
GSMA is a so-called "transport key", which encrypts the key material.
Let's introduce support for this in the CardKeyProvider (and
specifically, the CardKeyProviderCSV) and allow the user to specify
transport key material as command line options to pySim-shell.
Different transport keys can be used for different key materials, so
allow specification of keys on a CSV-column base.
The higher-level goal is to allow the CSV file not only to store
the ADM keys (like now), but also global platform key material for
establishing SCP towards various security domains in a given card.
Change-Id: I13146a799448d03c681dc868aaa31eb78b7821ff
... and populate the RuntimeState.identity['EID'] wit it, so other
[future] parts of the system can use it.
Let's also print the EID (if available) from the 'cardinfo' shell
command.
Change-Id: Idc2ea1d9263f39b3dff403e1535a5e6c4e88b26f
In the previous patch, we've introduced a new 'identities' dict as part
of the runtime state. Let's migrate our ICCID storage into it for
consistency.
Change-Id: Ibdcf9a7c4e7e445201640bce33b768bcc4460db1
Cards where no profile was detected don't have a logical channel, and
hence must use the raw APDU at all times.
Change-Id: I08e5d190bdb4e62ee808bfd77584cb3e0b85a8ae
Fixes: Change-Id Id0c364f772c31e11e8dfa21624d8685d253220d0
The 'apdu' command so far bypassed the logical channel and also
the recently-introduced support for secure channels. Let's change
that, at least by default. If somebody wants a raw APDU without
secure / logical channel processing, they may use the --raw option.
Change-Id: Id0c364f772c31e11e8dfa21624d8685d253220d0
This implementation of GlobalPlatform SCP02 currently only supports
C-MAC and C-ENC, but no R-MAC or R-ENC yet.
The patch also introduces the notion of having a SCP instance associated
with a SimCardCommands instance. It also adds the establish_scp0w and
release_scp shell commands to all GlobalPlatform Security Domains.
Change-Id: I56020382b9dfe8ba0f7c1c9f71eb1a9746bc5a27
Let's not have higher level code directly call the transports send_apdu*
methods. We do this as a precursor to introducing secure channel
support, where the secure channel driver would add MAC and/or encrypt
APDUs before they are sent to the transport.
Change-Id: I1b870140959aa8241cda2246e74576390123cb2d
If we are not 'equipped' as we could not detect any known applications
on the card, we used to only permit the 'apdu' command. However, we
should also permit the 'reset' command, as it also is something that's
possible with ever card, even of unknown types.
Change-Id: I23199da727973d7095ac18031f49e1e8423aa287
Some specs call it 'invalidated', others call it 'deactivated'. If the
user is unfamiliar with this, the error message about "invalidated"
might not be obvious enough; let's also mention 'deactivated' in the
message and explicitly mention that it needs to be activated before use.
Change-Id: I91488b0e7dc25a8970022b09e575485a4165eefa
When we traverse the file system, we may also end up selecting
applications (ADF), which do not support an USIM/ISIM like file system.
This will leave us without the ability to select the MF (or any other
file) again. The only way out is to select the ISIM or USIM application
again to get the access to the file system again.
Change-Id: Ia2fdd65f430c07acb1afdaf265d24c6928b654e0
Related: OS#5418
When sending raw APDUs, we access the scc (SimCardCommands) object via
the scc member in the lchan object. Unfortunately self.lchan will not be
populated when the rs (RuntimeState) object is missing. This is in
particular the case when no profile could be detected for the card,
which is a common situation when we boostrap an unprovisioned card.
So let's access the scc object through the card object. This is also
more logical since when we send raw APDUs we work below the level of
logical channels.
Change-Id: I6bbaebe7d7a2013f0ce558ca2da7d58f5e6d991a
Related: OS#6278
When there is an error on initialization (e.g. card not present), we
should not continue to execute a startup script that was passed with the
pySim-shell commandline. Instead we should print a message that the
startup script was ignored due to errors.
Related: OS#6271
Change-Id: I61329988e0e9021b5b0ef8e0819fb8e23cabf38b
The point of this is to move generic code out of pySim-shell.py,
paving the way for more/other executables using the full power of
our class model without having to reinvent the wheel.
Change-Id: Icf557ed3064ef613ed693ce28bd3514a97a938bd
Don't even send any non-decimal PIN values to the card, but reject
them when parsing the command arguments.
Change-Id: Icec1698851471af7f76f20201dcdcfcd48ddf365
before this patch:
pySIM-shell (00:MF)> echo foo bar baz
usage: echo [-h] string
echo: error: unrecognized arguments: bar baz
after this patch:
pySIM-shell (00:MF)> echo foo bar baz
foo bar baz
Change-Id: I1369bc3aa975865e3a8a574c132e469813a9f6b9
Let's make sure we don't even bother to ask the card to verify
anything as ADM1 pin which is not either a sequence of decimal digits
or an even number of hex digits (even number of bytes).
Change-Id: I4a193a3cf63462fad73d145ab1481070ddf767ca
Let's add a proper argparser instance for the 'verify_adm' command,
avoiding situations where the user types 'verif_adm --help' and then
--help is interpreted as the PIN value, removing one more attempt from
the failed ADM1 counter.
Let's use that opportunity to improve the documentation of the command.
Change-Id: I3321fae66a11efd00c53b66c7890fce84796e658
This just adds basic support for the ISD-R application and its
associated STORE DATA command which is used for the ES10x interfaces
between off-card entities and the on-card ISD-R.
Change-Id: Ieab37b083e25d3f36c20f6e9ed3e4bdfdd14a42a
Closes: OS#5637
Now that pySim-shell can switch between logical channels, let's state
the currently used logical channel in the prompt.
Change-Id: I45781a6fba205eeb4ac7f58d5cb642b7131bdd88
Related: OS#6230
We've already had the 'open_channel' and 'close_channel' commands,
which were sent to (and acknowledged by) the card. However,
those commands didn't affect the pySim-shell state, i.e. all
communication would still happen through the default channel '0'.
With this patch we introduce a 'switch_channel' command, using which
the user can determine which of the (previously opened) logical channels
shall be used by pySim-shell.
Change-Id: Ia76eb45c4925882ae6866e50b64d9610bd4d546d
Closes: OS#6230
This new approach will "fork" separate SimCardCommands instances
for each RuntimeLchan. Higher-layer code should now always use the
RuntimeLchan.scc rather than the RuntimeState.card._scc in order to
make sure commands use the correct logical channel.
Change-Id: I13e2e871f2afc2460d9fd1cd566de42267c7d389
Related: OS#6230
We already have the open channel and close_channel commands in
pySim-shell. They are sent to the card and acknowledged, respectively.
We also already do have code that can track multiple different logical
channels (the rs.lchan array). However, this is currently only used by
pySim-trace, and not by pySim-shell. Let's change that.
Change-Id: Idacee2dc57e8afe85c79bc85b259064e7f5b83a2
Related: OS#6230
An eUICC that has no active eSIM profile does not have an ICCID. (The
reason for this is that EF.ICCID is part of the eSIM profile).
Unfortunately pySim-shell insists on reading the ICCID from EF.ICCID on
startup in order to use it as a lookup key for verify_adm later.
To solve the problem, let's add a try/except block around the section
where EF.ICCID is read. In case of failure we set the ICCID to None,
Related: OS#5636
Change-Id: I8d18c5073946c5a6bb1f93be0ce692a599f46f8c
We currently catch any exceptions that may occur when the card reader is
initialized. Then we print the exception string or the exception type
when no string is available. However, a failure during the reader
initialization is usually a severe problem, so a traceback would provde
a lot of helpful information to debug the issue. So lets not catch any
exceptions at this level so that we get the full backtrace.
Related: OS#6210
Change-Id: I4c4807576fe63cf71a7d33b243a3f8fea0b7ff23
When an exception occurs while initializing or handling the card we
print a traceback, but we do not print any info that allows us to
identify the device that was involved when the exception occurred. Let's
include the device path or number in the error message before we print
the traceback.
In order to make it easier to print the device information, let's add a
__str__() method to all of our devices. This method shall return the
device number or path.
Related: OS#6210
Change-Id: I200463e692245da40ea6d5b609bfc0ca02d15bdb
When the try block in which we also call init_card() fails, there may be
no card object, so we must not pass the card object to PysimApp in the
except block. This is also no problem, PysimApp will run without the
card object until the user executes do_equip for a second attempt.
Related: OS#6210
Change-Id: I28195f442ce007f05f7610c882bbc4a6520a8ce6
When __main__ runs different variables get assigned. In particular opts,
scc, sl and ch. Those variables are available in any scope and
technically it is possible to access them. However, lets not do this
since it leads to confusion. Also, pylint will complain about those code
locations.
In pySim-shell.py
- Let's use the proper locations (sl and ch are stored in PysimApp.
- Scc can be assigned in init_card.
- In method walk, the use of the variable opts to call ection_df is wrong,
lets use **kwargs (see also usage of action_ef).
- The constructor of Cmd2ApduTracer has a parameter cmd2_app, but usese
the global variable app. Let's use cmd2_app instead.
In pySim-prog.py
- Do not use opts.num in find_row_in_csv_file, use num instead.
- Pass scc to process_card as parameter so that it won't access scc
in the global scope.
Change-Id: I7f09e9a6a6bfc658de75e86f7383ce73726f6666
Related: OS#6210
The commandline option -a, which does an ADM verification on startup,
does no longer work since the verify_adm method is no longer available
in the card base classes (cards.py). Let's use the verify_chv method
from SimCardCommands instead.
Related: RT#68294
Change-Id: Ic1e54d0e9e722d64b3fbeb044134044d47946f7c
The method read_iccid in class CardBase should be put back to
legacy/cards.py. The reason for this is that it falls in the same
category like read_imsi, read_ki, etc. We should not use those old
methods in future programs since we have a more modern infrastructure
(lchan) now.
Also pySim-shell.py is the only caller of this method now. It is not
used in any other place.
Related: RT#67094
Change-Id: Ied3ae6fd107992abcc1b5ea3edb0eb4bdcd2f892
When the command equip (do_equip) is executed, it accesses
self.rs.profile to see if there are any commands that need to be
unregistered before moving on with the card initialization.
However, it may be the case that no runtime state exists at this point.
This is in particular the case when the card is completely empty and
hence no profile is picked and no runtime state exists.
Change-Id: I0a8be66a69b630f1f2898b62dc752a8eb5275301
