2011-12-22 16:19:25 +01:00
|
|
|
/*
|
|
|
|
* \brief Test program for raising and handling region-manager faults
|
|
|
|
* \author Norman Feske
|
|
|
|
* \date 2008-09-24
|
|
|
|
*
|
|
|
|
* This program starts itself as child. When started, it first determines
|
2016-11-06 14:26:34 +01:00
|
|
|
* wheather it is parent or child by requesting a RM session. Because the
|
|
|
|
* program blocks all session-creation calls for the RM service, each program
|
|
|
|
* instance can determine its parent or child role by the checking the result
|
|
|
|
* of the session creation.
|
2011-12-22 16:19:25 +01:00
|
|
|
*/
|
|
|
|
|
|
|
|
/*
|
2017-02-20 13:23:52 +01:00
|
|
|
* Copyright (C) 2008-2017 Genode Labs GmbH
|
2011-12-22 16:19:25 +01:00
|
|
|
*
|
|
|
|
* This file is part of the Genode OS framework, which is distributed
|
2017-02-20 13:23:52 +01:00
|
|
|
* under the terms of the GNU Affero General Public License version 3.
|
2011-12-22 16:19:25 +01:00
|
|
|
*/
|
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
#include <base/component.h>
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
#include <base/log.h>
|
2011-12-22 16:19:25 +01:00
|
|
|
#include <base/child.h>
|
2015-05-05 08:50:16 +02:00
|
|
|
#include <rm_session/connection.h>
|
2017-01-30 11:35:12 +01:00
|
|
|
#include <base/attached_ram_dataspace.h>
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
using namespace Genode;
|
|
|
|
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
|
2011-12-22 16:19:25 +01:00
|
|
|
/***********
|
|
|
|
** Child **
|
|
|
|
***********/
|
|
|
|
|
|
|
|
enum { MANAGED_ADDR = 0x10000000 };
|
|
|
|
|
|
|
|
|
|
|
|
void read_at(addr_t addr)
|
|
|
|
{
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("perform read operation at ", Hex(addr));
|
2011-12-22 16:19:25 +01:00
|
|
|
int value = *(int *)addr;
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("read value ", Hex(value));
|
2011-12-22 16:19:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
void modify(addr_t addr)
|
|
|
|
{
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("modify memory at ", Hex(addr), " to ", Hex(++(*(int *)addr)));
|
2011-12-22 16:19:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
void main_child()
|
|
|
|
{
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("child role started");
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
/* perform illegal access */
|
|
|
|
read_at(MANAGED_ADDR);
|
|
|
|
|
|
|
|
while (true)
|
|
|
|
modify(MANAGED_ADDR);
|
|
|
|
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("--- child role of region-manager fault test finished ---");
|
2011-12-22 16:19:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
/************
|
|
|
|
** Parent **
|
|
|
|
************/
|
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
class Test_child_policy : public Child_policy
|
2011-12-22 16:19:25 +01:00
|
|
|
{
|
2016-11-06 14:26:34 +01:00
|
|
|
public:
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
typedef Registered<Genode::Parent_service> Parent_service;
|
|
|
|
typedef Registry<Parent_service> Parent_services;
|
2016-04-20 21:12:57 +02:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
private:
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
Env &_env;
|
|
|
|
Parent_services &_parent_services;
|
|
|
|
Signal_context_capability const _fault_handler_sigh;
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
public:
|
|
|
|
|
|
|
|
/**
|
|
|
|
* Constructor
|
|
|
|
*/
|
2016-11-06 14:26:34 +01:00
|
|
|
Test_child_policy(Env &env, Parent_services &parent_services,
|
|
|
|
Signal_context_capability fault_handler_sigh)
|
2011-12-22 16:19:25 +01:00
|
|
|
:
|
2016-11-06 14:26:34 +01:00
|
|
|
_env(env),
|
|
|
|
_parent_services(parent_services),
|
|
|
|
_fault_handler_sigh(fault_handler_sigh)
|
|
|
|
{ }
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
|
|
|
|
/****************************
|
|
|
|
** Child-policy interface **
|
|
|
|
****************************/
|
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
Name name() const override { return "rmchild"; }
|
|
|
|
|
|
|
|
Binary_name binary_name() const override { return "test-rm_fault"; }
|
2011-12-22 16:19:25 +01:00
|
|
|
|
Capability quota accounting and trading
This patch mirrors the accounting and trading scheme that Genode employs
for physical memory to the accounting of capability allocations.
Capability quotas must now be explicitly assigned to subsystems by
specifying a 'caps=<amount>' attribute to init's start nodes.
Analogously to RAM quotas, cap quotas can be traded between clients and
servers as part of the session protocol. The capability budget of each
component is maintained by the component's corresponding PD session at
core.
At the current stage, the accounting is applied to RPC capabilities,
signal-context capabilities, and dataspace capabilities. Capabilities
that are dynamically allocated via core's CPU and TRACE service are not
yet covered. Also, the capabilities allocated by resource multiplexers
outside of core (like nitpicker) must be accounted by the respective
servers, which is not covered yet.
If a component runs out of capabilities, core's PD service prints a
warning to the log. To observe the consumption of capabilities per
component in detail, the PD service is equipped with a diagnostic
mode, which can be enabled via the 'diag' attribute in the target
node of init's routing rules. E.g., the following route enables the
diagnostic mode for the PD session of the "timer" component:
<default-route>
<service name="PD" unscoped_label="timer">
<parent diag="yes"/>
</service>
...
</default-route>
For subsystems based on a sub-init instance, init can be configured
to report the capability-quota information of its subsystems by
adding the attribute 'child_caps="yes"' to init's '<report>'
config node. Init's own capability quota can be reported by adding
the attribute 'init_caps="yes"'.
Fixes #2398
2017-05-08 21:35:43 +02:00
|
|
|
Pd_session &ref_pd() override { return _env.pd(); }
|
|
|
|
Pd_session_capability ref_pd_cap() const override { return _env.pd_session_cap(); }
|
2016-11-06 14:26:34 +01:00
|
|
|
|
Capability quota accounting and trading
This patch mirrors the accounting and trading scheme that Genode employs
for physical memory to the accounting of capability allocations.
Capability quotas must now be explicitly assigned to subsystems by
specifying a 'caps=<amount>' attribute to init's start nodes.
Analogously to RAM quotas, cap quotas can be traded between clients and
servers as part of the session protocol. The capability budget of each
component is maintained by the component's corresponding PD session at
core.
At the current stage, the accounting is applied to RPC capabilities,
signal-context capabilities, and dataspace capabilities. Capabilities
that are dynamically allocated via core's CPU and TRACE service are not
yet covered. Also, the capabilities allocated by resource multiplexers
outside of core (like nitpicker) must be accounted by the respective
servers, which is not covered yet.
If a component runs out of capabilities, core's PD service prints a
warning to the log. To observe the consumption of capabilities per
component in detail, the PD service is equipped with a diagnostic
mode, which can be enabled via the 'diag' attribute in the target
node of init's routing rules. E.g., the following route enables the
diagnostic mode for the PD session of the "timer" component:
<default-route>
<service name="PD" unscoped_label="timer">
<parent diag="yes"/>
</service>
...
</default-route>
For subsystems based on a sub-init instance, init can be configured
to report the capability-quota information of its subsystems by
adding the attribute 'child_caps="yes"' to init's '<report>'
config node. Init's own capability quota can be reported by adding
the attribute 'init_caps="yes"'.
Fixes #2398
2017-05-08 21:35:43 +02:00
|
|
|
Ram_session &ref_ram() override { return _env.ram(); }
|
2016-11-06 14:26:34 +01:00
|
|
|
Ram_session_capability ref_ram_cap() const override { return _env.ram_session_cap(); }
|
|
|
|
|
|
|
|
void init(Ram_session &session, Ram_session_capability cap) override
|
|
|
|
{
|
|
|
|
enum { CHILD_QUOTA = 1*1024*1024 };
|
|
|
|
session.ref_account(_env.ram_session_cap());
|
2017-05-08 01:33:40 +02:00
|
|
|
_env.ram().transfer_quota(cap, Ram_quota{CHILD_QUOTA});
|
2016-11-06 14:26:34 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
void init(Pd_session &session, Pd_session_capability cap) override
|
2011-12-22 16:19:25 +01:00
|
|
|
{
|
Capability quota accounting and trading
This patch mirrors the accounting and trading scheme that Genode employs
for physical memory to the accounting of capability allocations.
Capability quotas must now be explicitly assigned to subsystems by
specifying a 'caps=<amount>' attribute to init's start nodes.
Analogously to RAM quotas, cap quotas can be traded between clients and
servers as part of the session protocol. The capability budget of each
component is maintained by the component's corresponding PD session at
core.
At the current stage, the accounting is applied to RPC capabilities,
signal-context capabilities, and dataspace capabilities. Capabilities
that are dynamically allocated via core's CPU and TRACE service are not
yet covered. Also, the capabilities allocated by resource multiplexers
outside of core (like nitpicker) must be accounted by the respective
servers, which is not covered yet.
If a component runs out of capabilities, core's PD service prints a
warning to the log. To observe the consumption of capabilities per
component in detail, the PD service is equipped with a diagnostic
mode, which can be enabled via the 'diag' attribute in the target
node of init's routing rules. E.g., the following route enables the
diagnostic mode for the PD session of the "timer" component:
<default-route>
<service name="PD" unscoped_label="timer">
<parent diag="yes"/>
</service>
...
</default-route>
For subsystems based on a sub-init instance, init can be configured
to report the capability-quota information of its subsystems by
adding the attribute 'child_caps="yes"' to init's '<report>'
config node. Init's own capability quota can be reported by adding
the attribute 'init_caps="yes"'.
Fixes #2398
2017-05-08 21:35:43 +02:00
|
|
|
session.ref_account(_env.pd_session_cap());
|
|
|
|
_env.pd().transfer_quota(cap, Cap_quota{20});
|
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
Region_map_client address_space(session.address_space());
|
|
|
|
address_space.fault_handler(_fault_handler_sigh);
|
|
|
|
}
|
|
|
|
|
|
|
|
Service &resolve_session_request(Service::Name const &service_name,
|
|
|
|
Session_state::Args const &args) override
|
|
|
|
{
|
|
|
|
Service *service = nullptr;
|
|
|
|
_parent_services.for_each([&] (Service &s) {
|
|
|
|
if (!service && service_name == s.name())
|
|
|
|
service = &s; });
|
|
|
|
|
|
|
|
if (!service)
|
|
|
|
throw Parent::Service_denied();
|
|
|
|
|
|
|
|
return *service;
|
2011-12-22 16:19:25 +01:00
|
|
|
}
|
|
|
|
};
|
|
|
|
|
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
struct Main_parent
|
2011-12-22 16:19:25 +01:00
|
|
|
{
|
2016-11-06 14:26:34 +01:00
|
|
|
Env &_env;
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
Signal_handler<Main_parent> _fault_handler {
|
|
|
|
_env.ep(), *this, &Main_parent::_handle_fault };
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
Heap _heap { _env.ram(), _env.rm() };
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
/* parent services */
|
|
|
|
struct Parent_services : Test_child_policy::Parent_services
|
|
|
|
{
|
|
|
|
Allocator &alloc;
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
Parent_services(Allocator &alloc) : alloc(alloc)
|
|
|
|
{
|
|
|
|
static const char *names[] = {
|
|
|
|
"RAM", "PD", "CPU", "ROM", "LOG", 0 };
|
|
|
|
for (unsigned i = 0; names[i]; i++)
|
|
|
|
new (alloc) Test_child_policy::Parent_service(*this, names[i]);
|
|
|
|
}
|
|
|
|
|
|
|
|
~Parent_services()
|
|
|
|
{
|
|
|
|
for_each([&] (Test_child_policy::Parent_service &s) { destroy(alloc, &s); });
|
|
|
|
}
|
|
|
|
} _parent_services { _heap };
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
/* create child */
|
2016-11-06 14:26:34 +01:00
|
|
|
Test_child_policy _child_policy { _env, _parent_services, _fault_handler };
|
|
|
|
|
|
|
|
Child _child { _env.rm(), _env.ep().rpc_ep(), _child_policy };
|
|
|
|
|
|
|
|
Region_map_client _address_space { _child.pd().address_space() };
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
/* dataspace used for creating shared memory between parent and child */
|
|
|
|
Attached_ram_dataspace _ds { _env.ram(), _env.rm(), 4096 };
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
unsigned _fault_cnt = 0;
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
long volatile &_child_value() { return *_ds.local_addr<long volatile>(); }
|
|
|
|
|
|
|
|
void _handle_fault()
|
|
|
|
{
|
|
|
|
if (_fault_cnt++ == 4) {
|
|
|
|
log("--- parent role of region-manager fault test finished ---");
|
|
|
|
_env.parent().exit(0);
|
|
|
|
}
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
log("received region-map fault signal, request fault state");
|
|
|
|
|
|
|
|
Region_map::State state = _address_space.state();
|
2011-12-22 16:19:25 +01:00
|
|
|
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
char const *state_name =
|
|
|
|
state.type == Region_map::State::READ_FAULT ? "READ_FAULT" :
|
|
|
|
state.type == Region_map::State::WRITE_FAULT ? "WRITE_FAULT" :
|
|
|
|
state.type == Region_map::State::EXEC_FAULT ? "EXEC_FAULT" : "READY";
|
|
|
|
|
|
|
|
log("rm session state is ", state_name, ", pf_addr=", Hex(state.addr));
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
/* ignore spuriuous fault signal */
|
2016-04-15 15:19:22 +02:00
|
|
|
if (state.type == Region_map::State::READY) {
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("ignoring spurious fault signal");
|
2016-11-06 14:26:34 +01:00
|
|
|
return;
|
2011-12-22 16:19:25 +01:00
|
|
|
}
|
|
|
|
|
|
|
|
addr_t child_virt_addr = state.addr & ~(4096 - 1);
|
|
|
|
|
|
|
|
/* allocate dataspace to resolve the fault */
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("attach dataspace to the child at ", Hex(child_virt_addr));
|
2016-11-06 14:26:34 +01:00
|
|
|
_child_value() = 0x1234;
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
_address_space.attach_at(_ds.cap(), child_virt_addr);
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
/* poll until our child modifies the dataspace content */
|
|
|
|
while (_child_value() == 0x1234);
|
2011-12-22 16:19:25 +01:00
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
log("child modified dataspace content, new value is ",
|
|
|
|
Hex(_child_value()));
|
2011-12-22 16:19:25 +01:00
|
|
|
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("revoke dataspace from child");
|
2016-11-06 14:26:34 +01:00
|
|
|
_address_space.detach((void *)child_virt_addr);
|
2011-12-22 16:19:25 +01:00
|
|
|
}
|
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
Main_parent(Env &env) : _env(env) { }
|
|
|
|
};
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
|
2016-11-06 14:26:34 +01:00
|
|
|
void Component::construct(Env &env)
|
2011-12-22 16:19:25 +01:00
|
|
|
{
|
base: avoid use of deprecated base/printf.h
Besides adapting the components to the use of base/log.h, the patch
cleans up a few base headers, i.e., it removes unused includes from
root/component.h, specifically base/heap.h and
ram_session/ram_session.h. Hence, components that relied on the implicit
inclusion of those headers have to manually include those headers now.
While adjusting the log messages, I repeatedly stumbled over the problem
that printing char * arguments is ambiguous. It is unclear whether to
print the argument as pointer or null-terminated string. To overcome
this problem, the patch introduces a new type 'Cstring' that allows the
caller to express that the argument should be handled as null-terminated
string. As a nice side effect, with this type in place, the optional len
argument of the 'String' class could be removed. Instead of supplying a
pair of (char const *, size_t), the constructor accepts a 'Cstring'.
This, in turn, clears the way let the 'String' constructor use the new
output mechanism to assemble a string from multiple arguments (and
thereby getting rid of snprintf within Genode in the near future).
To enforce the explicit resolution of the char * ambiguity, the 'char *'
overload of the 'print' function is marked as deleted.
Issue #1987
2016-07-13 19:07:09 +02:00
|
|
|
log("--- region-manager fault test ---");
|
2011-12-22 16:19:25 +01:00
|
|
|
|
|
|
|
try {
|
2016-11-06 14:26:34 +01:00
|
|
|
/*
|
|
|
|
* Distinguish parent from child by requesting an service that is only
|
|
|
|
* available to the parent.
|
|
|
|
*/
|
2017-01-05 16:19:16 +01:00
|
|
|
Rm_connection rm(env);
|
2016-11-06 14:26:34 +01:00
|
|
|
static Main_parent parent(env);
|
|
|
|
log("-- parent role started --");
|
|
|
|
}
|
|
|
|
catch (Parent::Service_denied) {
|
2011-12-22 16:19:25 +01:00
|
|
|
main_child();
|
|
|
|
}
|
|
|
|
}
|