class documentation

class Context(object):

Entry point object for data related to a context.

A context represents the full state of the VM at a point in time. You can access contexts before or after a transition:

>>> ... -> Context Before #N -> Transition #N -> Context After #N ->

A context can usually be referenced as both a context before and a context after:

>>> ... -> Transition #N-1 -> Context After #N-1 / Before #N -> Transition #N -> ...

Except at the trace's begin and end:

>>> Context Before #0 -> Transition #0 -> ... -> Transition #M -> Context After #M

Warnings

This object is not meant to be constructed directly. Use Trace.context_before, Trace.context_after, Transition.context_before or Transition.context_after instead.

Examples

>>> # From a transition
>>> reven_server.trace.transition(id).context_before()
>>> reven_server.trace.transition(id).context_after()

>>> # From a trace directly
>>> reven_server.trace.context_before(transition_id)
>>> reven_server.trace.context_after(transition_id)

>>> # From a context
>>> ctx = reven_server.trace.context_before(transition_id)
>>> next_ctx = ctx + 1
>>> prev_ctx = ctx - 1
>>> other_ctx = ctx + 10000

Method	`__init__`	Undocumented
Method	`transition_before`	Get Transition that led to this context
Method	`transition_after`	Get Transition that will execute right after this
Property	`ossi`	Property: OSSI information related to the context
Property	`stack`	Property: The active stack linked to the context.
Method	`is64b`	Undocumented
Method	`physical_memory_regions`	Get all the physical memory regions that are accessible.
Method	`read`	read(register [, raw=False]) -> read bytes from a register read(source, size[, raw=False]) -> read bytes from a register or a memory read(source, type) -> read bytes from a register or a memory
Method	`deref`	Recursively dereferences a pointer and read the value read as a pointing type.
Method	`find_register_change`	Find the next/previous context in which the content of the register is modified.
Method	`search_in_memory`	No summary
Method	`__str__`	Undocumented
Method	`__repr__`	Undocumented
Method	`format_as_html`	This method gets an html formatting string representation for this class instance.
Method	`__eq__`	Undocumented
Method	`__ne__`	Undocumented
Method	`__lt__`	Undocumented
Method	`__le__`	Undocumented
Method	`__gt__`	Undocumented
Method	`__ge__`	Undocumented
Method	`__add__`	Undocumented
Method	`__sub__`	Undocumented
Instance Variable	`__trace`	Undocumented
Instance Variable	`_data_source`	Undocumented
Instance Variable	`_ossi_data_source`	Undocumented
Instance Variable	`_id`	Undocumented
Instance Variable	`_cpu_context`	Undocumented
Property	`_trace`	Undocumented
Static Method	`_next_stop_point`	Undocumented
Method	`_read_size`	Undocumented
Method	`_read_as`	Undocumented
Method	`_read_raw`	Undocumented
Method	`_read_st_register`	Undocumented
Method	`_read_register`	Undocumented
Method	`_repr_html_`	Representation used by Jupyter Notebook when an instance of this class is displayed in a cell.
Property	`_rvn`	Undocumented

def __init__(self, trace, _data_source, _ossi_data_source, context_id):

Undocumented

__trace =

Undocumented

_data_source =

Undocumented

_ossi_data_source =

Undocumented

_id =

Undocumented

_cpu_context =

Undocumented

@property
_trace =

Undocumented

def transition_before(self):

Get Transition that led to this context

>>> Transition Before -> This context

Information

Returns	A `Transition`.
Raises	IndexError	if called on the context before the first transition

def transition_after(self):

Get Transition that will execute right after this

>>> This context -> Transition After

Information

Returns	A `Transition`.
Raises	IndexError	if called on the context after the last transition

@property
ossi =

Property: OSSI information related to the context

Examples

>>> # the current symbol
>>> print(context.ossi.location())
'ntoskrnl!KiIsrLinkage+0x10'

Information

Returns A reven2.ossi.OssiContext.

@property
stack =

Property: The active stack linked to the context.

Examples

>>> # the active stack
>>> print(context.stack.backtrace)
[0] #2753249448 - ntoskrnl!KeZeroPages
[1] #2753249199 - ntoskrnl!MiZeroPhysicalPage
[2] #2753248836 - ntoskrnl!MiResolveDemandZeroFault
[3] #2753248607 - ntoskrnl!MmAccessFault
[4] #2753248552 - ntoskrnl!KiPageFault
[5] ??? - ntoskrnl!KiSystemCall64+0x15

Information

Returns A reven2.stack.Stack.

def is64b(self):

Undocumented

def physical_memory_regions(self):

Get all the physical memory regions that are accessible.

These regions are assumed to not change during the entire trace.

Any physical read inside these regions is guaranteed to succeed.

Examples

>>> for (start, size) in ctx.physical_memory_regions():
>>>     print("{} - {}".format(start, start + size))
phy:0x0 - phy:0x80000000
phy:0xfd000000 - phy:0xfe000000

Informations

Returns A generator of tuples composed of the start physical address of a region and its size.

def read(self, source, *args, **kwargs):

read(register [, raw=False]) -> read bytes from a register read(source, size[, raw=False]) -> read bytes from a register or a memory read(source, type) -> read bytes from a register or a memory

If 'size' in [1, 2, 4, 8], the bytes are interpreted as an unsigned int. If 'raw' is True, a raw buffer is returned even if 'size' is in [1, 2, 4, 8]. If 'type' is provided (a reven2.types.Type), the bytes are interpreted as the given 'type'. In all other cases, a raw buffer is returned.

If 'source' is a virtual address, then [source; length] must be mapped in the current context, with `length == size` if 'size' was provided, or `length == type.size_bytes(self)` if `type` was provided

If 'source' is a physical address, then [source; length] must be in the available physical memory (see Context.physical_memory_regions), with `length == size` if 'size' was provided, or `length == type.size_bytes(self)` if `type` was provided

Examples

Read rax without size:

>>> hex(context.read(arch.x64.rax))
'0xffffc000e9997ac0'

>>> Get a raw buffer
>>> context.read(arch.x64.rax, raw=True)
bytearray(b'\xc0z\x99\xe9\x00\xc0\xff\xff')

Read rax with size:

>>> # size not in [1, 2, 4, 8]
>>> context.read(arch.x64.rax, 6)
bytearray(b'\xc0z\x99\xe9\x00\xc0')

>>> # size in [1, 2, 4, 8]
>>> hex(context.read(arch.x64.rax, 4))
'0xe9997ac0'

>>> # Get a raw buffer
>>> hex(context.read(arch.x64.rax, 4, raw=True))
bytearray(b'\xc0z\x99\xe9')

Read rax as a U64 integer:

>>> hex(context.read(arch.x64.rax, types.U64))
'0xffffc000e9997ac0'

Read rax as two U8 integers:

>>> context.read(arch.x64.rax, types.Array(types.U16, 2))
ListContainer([0xe999, 0x7ac0])

Read a buffer in memory at address ds:0x1234:

>>> address = address.LogicalAddress(0x1234)

>>> # size not in [1, 2, 4, 8]
>>> context.read(address, 6)
bytearray(b'\xc0z\x99\xe9\x00\xc0')

>>> # size in [1, 2, 4, 8]
>>> hex(context.read(address, 8))
'0xffffc000e9997ac0'

>>> # Get a raw buffer
>>> context.read(address, 8, raw=True)
bytearray(b'\xc0z\x99\xe9\x00\xc0\xff\xff')

Read a string in memory at address ds:0x1234:

>>> address = address.LogicalAddress(0x1234)
>>> WString1000 = types.CString(encoding=types.Encoding.Utf16, max_character_count=1000)
>>> context.read(address, WString1000)
'Hello World'

Information

Parameters	source	From where the bytes are read. Can be an address (virtual or physical) or a `Register`.
	args	Undocumented
	kwargs	Undocumented
	type	A `reven2.types.Type`. See the `reven2.types` package documentation for more information on types.
	size	The byte size to read. If no 'size' and 'source' is a `reven2.arch.register.Register`, then the `Register.size_bytes` is used.
	raw	Force the return type to be a raw buffer of 'size' bytes.
Raises	TypeError	if 'source' is not supported.
	TypeError	if 'type' is not a `Type`.
	ValueError	if 'source' is an `address` and 'size' or 'type' are not provided.
	ValueError	if 'source' is a `Register` and 'size' is bigger than the register byte size.
	RuntimeError	if 'source' is an unavailable register.
	RuntimeError	if 'source' is a virtual address and is not mapped in the current context.
	RuntimeError	if 'source' is a physical address and is outside the physical memory.

def deref(self, source, ptr_ty):

Recursively dereferences a pointer and read the value read as a pointing type.

This method allows to directly returns the pointed-to value when passed a pointer type and a source from which read the pointer address.

A (simplified) desugaring of this method can be thought of like the following:

>>> addr = self.read(source, ptr_ty)
>>> return self.read(addr, ptr_ty.elem_type)

Note, however that the actual implementation will recursively remove all levels of indirection, e.g. if the type is Pointer(Pointer(U64)), then the implementation will do the following:

>>> addr1 = self.read(source, Pointer(Pointer(U64)))
>>> addr2 = self.read(addr1, Pointer(U64))
>>> return self.read(addr2, U64)

Example

Reading the stack pointer:

>>> hex(context.deref(reven2.arch.x64.rsp, types.Pointer(types.USize)))
'0xfffff801653ff072L'

Looks like an address, let's try to dereference twice:

>>> hex(context.deref(reven2.arch.x64.rsp, types.Pointer(types.Pointer(types.USize))))
'0x7c894828247c894c'

Doesn't look like an address anymore, show raw:

>>> [hex(x) for x in context.deref(reven2.arch.x64.rsp, types.Pointer(types.Pointer(types.RawBuffer(8))))]
['0x4c', '0x89', '0x7c', '0x24', '0x28', '0x48', '0x89', '0x7c']

Information

Parameters	source	The source where to read the address of the pointer. Can be an address (virtual or physical) or a register.
	ptr_ty	A `reven2.types.Pointer`.
Returns	The data interpreted as the requested type. If the pointer points to an unmapped memory, `None` is returned.
Raises	TypeError	if 'source' is not supported.
	TypeError	if 'ptr_ty' is not a `Pointer`.
	ValueError	if 'source' is a `Register` and there is not enough bytes to read the pointer address in it.
	RuntimeError	if 'source' is an unavailable register.

def find_register_change(self, register, is_forward=True, fetch_count=10000):

Find the next/previous context in which the content of the register is modified.

Information

Parameters	register	`reven2.arch.register.Register`
	is_forward	If `True` return next context. Otherwise return previous context.
	fetch_count	Technical parameter indicating how many contexts the server should handle per query. Modifying this parameter allows to fine-tune performance: A large value leads to better performance but may cause the script to block longer while waiting for the server response without being interruptible. A small value leads to a lot of server requests and so reduces performance but ensures the script will not block. The default value should provide sufficient performance in most cases. NOTE: Modifying this parameter does not modify the results of the query
Returns	A `reven2.trace.Context` or `None` if no context was found.
Raises	ValueError	if `fetch_count` is not positive.

@staticmethod
def _next_stop_point(is_forward, start, fetch_count, trans_count):

Undocumented

def search_in_memory(self, pattern, address, size, is_forward=True):

Iterate the search range looking for the specified pattern. Search range is:

[address, address + size[ in forward direction
[address - size, address[ in backward direction

This method returns a generator of addresses corresponding to the addresses where the searched pattern matches.

Information

Parameters	pattern	A `str` or `bytearray`. The pattern to look for in memory. Note: `str` pattern is converted to bytearray using ascci encoding.
	address	a class from `reven2.address` the address where the search will be started. an `Integer` represents the offset of the address where the search will be started.
	size	an `Integer` representing the size, in bytes, of the search range.
	is_forward	`bool`, `True` to search in forward direction and `False` to search in backward direction
Returns	a generator of addresses corresponding to the addresses where the searched pattern matches.
Raises	TypeError	if address is not a `Integer` or one of the address classes on `reven2.address`.
	RunTimeError	If the address is a virtual address that is not aligned to its corresponding physical address.

def _read_size(self, source, size, raw):

Undocumented

def _read_as(self, source, ty):

Undocumented

def _read_raw(self, source, size):

Undocumented

def _read_st_register(self, register, size):

Undocumented

def _read_register(self, register, size):

Undocumented

def __str__(self):

Undocumented

def __repr__(self):

Undocumented

def format_as_html(self):

This method gets an html formatting string representation for this class instance.

Information

Returns String

def _repr_html_(self):

Representation used by Jupyter Notebook when an instance of this class is displayed in a cell.

def __eq__(self, other):

Undocumented

def __ne__(self, other):

Undocumented

def __lt__(self, other):

Undocumented

def __le__(self, other):

Undocumented

def __gt__(self, other):

Undocumented

def __ge__(self, other):

Undocumented

def __add__(self, other):

Undocumented

def __sub__(self, other):

Undocumented

@property
_rvn =

Undocumented

API Documentation for reven2, generated by pydoctor 21.2.2 at 2021-04-08 12:32:07.

reven2.trace.Context

Warnings

Examples

Information

Information

Examples

Information

Examples

Information

Examples

Informations

Examples

Information

Example

Information

Information

Information

Information

`reven2.trace.Context`