This patch adds the trace-logger utility to the default set of packages
along with an optional launcher. With this change, only two steps are
needed to use Genode's tracing mechanism with Sculpt:
- Add 'trace_logger' to the 'launcher:' list of the .sculpt file
- Either manually select the 'trace_logger' from the '+' menu,
or add the following entry to the deploy configuration:
<start name="trace_logger"/>
By default, the trace logger is configured to trace all threads
executed in the runtime subsystem and to print a report every 10
seconds. This default policy can be refined in the launcher's <config>
node. Note that the trace logger does not respond to configuration
changes during runtime. Changes come into effect not before restarting
the component.
Issue #4448
This patch changes the output format of the trace logger to become
better suitable for human consumption. For example, when instrumenting
the VFS server in Sculpt using the GENODE_TRACE_TSC utility, the
trace logger now generates tabular output as follows.
Report 4
PD "init -> runtime -> arch_vbox6 -> vbox -> " ----------------
Thread "vCPU" at (0,0) total:12909024 recent:989229
Thread "vCPU" at (1,0) total:5643234 recent:786437
PD "init -> runtime -> ahci-0.fs" -----------------------------
Thread "ahci-0.fs" at (0,0) total:910497 recent:6335
Thread "ep" at (0,0) total:0 recent:0
71919692932: TSC process_packets: 8005M (4998 calls, last 4932K)
71921558516: TSC process_packets: 8006M (4999 calls, last 1596K)
71922760220: TSC process_packets: 8007M (5000 calls, last 1006K)
71929853586: TSC process_packets: 8009M (5001 calls, last 1840K)
71931315246: TSC process_packets: 8011M (5002 calls, last 1253K)
72127999920: TSC process_packets: 8016M (5003 calls, last 5606K)
72129568198: TSC process_packets: 8018M (5004 calls, last 1345K)
77161908178: TSC process_packets: 8029M (5005 calls, last 11349K)
77643225736: TSC process_packets: 8029M (5006 calls, last 217K)
89422100594: TSC process_packets: 8035M (5007 calls, last 5656K)
89422123632: TSC process_packets: 8035M (5008 calls, last 1342)
Thread "signal handler" at (0,0) total:36329 recent:3001
Thread "signal_proxy" at (0,0) total:51838 recent:13099
Thread "pdaemon" at (0,0) total:97184 recent:332
Thread "vdrain" at (0,0) total:1266 recent:286
Thread "vrele" at (0,0) total:1904 recent:516
PD "init -> runtime -> nic_drv" -------------------------------
Thread "nic_drv" at (0,0) total:34044 recent:897
Thread "signal handler" at (0,0) total:369 recent:142
...
Subjects that belong to the same PD are grouped together. The formerly
optional affinity and activity options have been removed. Those
information are now unconditionally displayed. The trace entries
belonging to a thread appear as slightly indented.
The patch also updates the coding style, avoiding excessively long
lines.
Issue #4448
This patch reduces repetitive log output by omitting inactive trace
subjects from the log output. The information about all subjects can
still be dumped by setting 'verbose="yes"'.
Issue #4448
This patch splits the creation and updating of monitor objects into two
stages. The creation of a monitor object changes the state of the
associated trace subject. The patch ensures that the new state is
captured by the update of the monitor object.
Issue #4448
This patch makes the trace-subject state as reflected to the trace
monitor more accurate.
Until now, a subject could be in UNTRACED or TRACED state. In reality,
however, there exists an intermediate state after the trace monitor
called 'trace' for the subject but before the subject locally activated
the tracing (done when passing a trace point). This intermediate state
was reflected as UNTRACED. Consequently, threads that never pass a trace
point (e.g., just waiting for I/O) would remain to appear as UNTRACED
even after enabling its tracing by the trace monitor. This is confusing.
This patch replaces the former UNTRACED and TRACED states by three
distinct states:
UNATTACHED prior any call of 'trace'
ATTACHED after a trace monitor called 'trace'
but before the tracing is active
TRACE tracing is active
Fixes#4447
The new macros GENODE_TRACE_TSC and GENODE_TRACE_TSC_NAMED complement
the existing GENODE_LOG_TSC and GENODE_LOG_TSC_NAMED macros to simplify
TSC measurements at a low overhead of the trace mechanism.
The utilities of the new util/formatted_output.h header complement the
existing base/output.h with the text-formatting support needed to
produce tabular output.
Fixes#4449
First, the endpoint update has to comply with the current alternate
settings of all interface, which are stored in USBDevice::altsetting[]
(one value per interface). Second, a SET_INTERFACE control request via
Packet_type::ALT_SETTING must update USBDevice::altsetting for the
interface.
Now, USB devices with multi-setting interface like the Joulescope JS110
with mixed bulk/isochronous endpoints are supported.
This commit adjusts the value such that USB sessions requested by
VirtualBox6 on Sculpt OS can get established on the first try without
invoking the session-retry mechanism. This reduces the number of
diagnostic log messages like:
Error: Insufficient 'ram_quota',got 6296372 need 6297928
The value of 50% as assigned by commit "sculpt: add basic support for
i.MX8 Quad EVK" leaves too little room for other components of the
runtime subsystem. With the adjustements of commit "sculpt: assign CPU
quotas" the sum exeedcs 100%. Hence this commit tunes down the value to
sensible 10%.
The existing assignment of CPU quotas did not anticipate the dynamic
reconfiguration of init. It merely tracked the available CPU quota by
deducing the consumed amount from a global variable but never
replenished the value. This worked for static scenarios but failed in
situations where components are dynamically re-started.
So far this deficiency remained detected because CPU quotas were not
used in highly dynamic systems like Sculpt OS. However, this has
recently changed by commit "sculpt: assign CPU quotas".
The patch improves the accounting by mirroring the existing handling of
RAM and cap quotas. Note that the CPU-quota accounting is still rather
limited. In particular the dynamic rebalancing is not yet supported.
Issue #4445
.SHELLFLAGS is extended by option pipefail to make pipes fail if any pipe
element fails. As .SHELLFLAGS is exported into sub-make instances it
must be unexported before calling third-party build systems recursively.
With the consolidation of the file-system session's signal handlers
implemented by commit "file_system_session: merge ack and submit sigh",
we can now change the VFS server to produce batches of acknowledgements
before explicitly waking up the client. (in contrast to the traditional
'acknowledge_packet', the new 'try_ack_packet' triggers no signal)
Issue #4388
Split the trace buffer into two partitions in order to prevent overwriting
of entries when the consumer is too slow. See file comment in buffer.h.
genodelabs/genode#4434
This commit simplifies the current implementation by overloading the
length field with a padding indicator in addition to the zero-length
head entry. This simplifies the iteration semantics as it eliminates
the need for determining whether a zero-length entries is the actual
head of the buffer or a padding at the buffer end.
genodelabs/genode#4434
To support device-less protocol-stacks only ports, we can use a
Virt I/O Linux kernel flavor, and export the lx_kit/lx_emul parts
not depending on platform API and devices.
Ref #4397
* Drivers have to use lx_kit/memory_dma.cc
* Protocol-stacks use lx_kit/memory_non_dma.cc
* Moreover the device-dependent lx_emul parts get removed
from the common lx_emul import rules
Fix#4443
* Unifies the declaration of callbacks which manage driver/client
shared dataspaces
* Move the Linux driver-specific callback implementation to the
lx_emul library from the PC's USB host driver
Fixgenodelabs/genode#4439
The logic got accidentally reversed during the transition from the
legacy USB driver.
Since the function drops error details, this patch adds a diagnostic
message with the error code as returned by the Linux driver.
This patch declares that the pseudo target virtualbox6/services/ does
not produce any build artifact. This allows the adding of virtualbox6
to Sculpt's 'build:' list.
The run script uses core, ld.lib.so, and init from depot packages, thus
these cannot be integrated in the image explicitly from the build
directory.
Also, removed special build of report_rom, which is not subject of the
test.
Fixes#4437
* The problem occured only on OKL4, seL4, and L4/Fiasco
* The test used to continue submitting packets regardless of how many were in
flight
* This commit limits the in-flight packets at the Uplink/Nic connections to 40
each
Ref #4419
This commit changes the touch-click filter to adhere the following
sequence:
touch (physical)
press (artifically generated)
release (artificially generated)
release_touch (physical)
This order is important because nitpicker's focus handling takes
press/release events into account. If the release-touch event appears
before the release event, nitpicker subsumes the release-touch event
to the sequence that started with the press event, instead of handling
it as a free-standing event.
Issue #4332
By assigning CPU budgets, base-hw's priorities come into effect.
Even though the chosen values are rather guessed than informed, they
yield a visibly improved responsiveness on the Pinephone.
When committing a new entry, the buffer wrapped if the last entry fit
perfectly into the buffer. Otherwise, the length field of the next entry
was set to 0 to mark the new head. Yet, if there was still some padding but not
enough to hold the length field of another entry, we ended up with a
headless buffer.
genodelabs/genode#4430
Since the head of the buffer is marked by a zero-length entry, we must
only write the length field if a new head was set. Otherwise, the
consumer might already read the new entry and not find the new head as a stop
condition.
genodelabs/genode#4430
When a device got already acquired by a platform session client, do not
return the same capability again, even if the same platform session
client requested it. Therefore, make doubtful behaviour of client components
visible, and do not have to struggle with Platform::Device instances
instantiated multiple times, which care of the lifetime of the device
capability internally.
Fixgenodelabs/genode#4379
Genode code already expects MMU to be disabled when starting the
kernel. It is enabled eventually in Bootstrap::Platform::enable_mmu,
after setting up translation tables. Unfortunately nothing ensures
this is actually the case. If MMU happens to be enabled when entering
the kernel things go downhill pretty fast after we start messing with
TTBR.
This patch ensures MMU is disabled for EL1, EL2, EL3 dependent on the
exception level of the CPU core, which is entering the kernel.
This should allow base-hw to start correctly on Quartz64 A board.
The `lookup_translation` function got introduced and is used only in the
context to proof whether a cache maintainance function can be executed
safely by the kernel. Unfortunately, it did not checked write permissions,
which can lead to permission faults. This commit restricts the lookup
function to only succeed when the target page is writeable. Consequently,
the lookup function gets renamed to `lookup_rw_translation`.
Fixgenodelabs/genode#4348
* Switch from the legacy usb_host driver to the new PC version
in recipes and automated tests
* Update documentation snippets
* Remove outdated, unused usb_rndis run-script
Fixgenodelabs/genode#4416
Do not only report devices, but the current configuration of the driver too,
as long as the `report` node in the configuration states it.
Ref genodelabs/genode#4416
To stay consistent with the configuration of the legacy usb_host_drv
and other components as well, do not report USB devices by default,
but when the following XML node is set within the component's
configuration:
<report devices="yes"/>
Ref genodelabs/genode#4416
The new usb_host driver for PC by default needs more RAM quota in contrast
to the old one. To minor warnings and repeated attempts to open a USB
session from a client, this commit increases the default RAM quota by one
page.
Ref genodelabs/genode#4416
Adds a function to the USB part of the Genode's C API, to enable
usb_host drivers to acknowledge USB request in client's packet buffer
although they are not assigned to an USB device. The requests are
marked with a "no device" error.
This commit fixes a regression originally solved in genodelabs/genode#4149
Ref genodelabs/genode#4416
XML allows attribute values like <node attr="\"/>. The XML parser
wrongly reflects this case as 'Invalid_syntax'. This behavior stems from
the implicit use of the 'end_of_quote' function, which considers the
sequence of '\"' as a quoted '"' rather than the end of a quoted string.
The patch solves this problem by making the 'end_of_quote' part of
the tokenizer's scanner policy.
The patch removes the 'end_of_quote' function from 'util/string.h'
because it is not universal, and to avoid the ambiguity with
'SCANNER_POLICY::end_of_quote'.
Fixes#4431
When a PD owns the right to ask for a RAM dataspace's `dma_addr` it is
concurrently constrained to use allocations of a specific physical RAM area.
This commit further limits this area by removing RAM page frame zero.
Otherwise the return value of `dma_addr` for such a dataspace would be
erroneously interpreted as a fault, because zero is currently the error
return value of `dma_addr`.
Fixgenodelabs/genode#4428
This patch cuts the superfluous dependency of abi.so files from the
library dependencies of the corresponding lib.so file. ABIs depend
only on symbol files.
Prior this patch, the second step of the following sequence would
wrongly re-build the abi.so file.
arm_v6$ make init
arm_v6$ make init KERNEL=hw
As the KERNEL argument does not affect the ABI, the abi.so should
better not be created twice.
Issue #4408
The Single_file_system now forwards the
`File_io_service::notify_read_ready` method to the
handle as it already did for most of the other methods.
genodelabs/genode#4394
Do not set '_mode' per default to MANANGED. Check if a manual config ROM
is present during construction and set '_mode' to MANUAL in case it is.
issue #4369
'generate(Xml_node node)', as used by the Sculpt manager, calls this
function instead of the lambda version. The 'report' function of the
'Genode::Reporter' does not throw an exception in case there is not
enough backing storage for the 'generate' request. Therefore, we have to
check this condition in a loop and call '_increase_report_buffer' in
case size limits are reached.
Patch by Norman Feske.
issue #4369
Genode linker does not support .gnu.hash tables so they will never be
used. Tell the linker not to bother producing them. This should reduce
the size of Genode ELF files a tiny bit without loosing anything
important in the process.
Fixes#4423
According to ARM Cortex-A55 Core Technical Reference Manual r1p0 the
lowest 8 bits (Aff0) of MPIDR register represent thread IDs within a
multi-threaded core. The actual core identification bits are in Aff1.
This layout can be identified by checking the MT bit of MPIDR register.
Basically, if MT=1 core id is in Aff1, if MT=0 core id is in Aff0.
Without this change Genode will identify all CPU cores on A55 as primary
(0) core.
Its worth to mention that Cortex-A55 by itself is not a multi-threaded
CPU. Aff0 values are always expected to be 0 for pure A55 cores. A55
cores can however be paired with cores that are multi-threaded. To
support such big.LITTLE CPUs in Genode we'd probably need to add a
different mechanism for mapping MPIDR values to logical, contignous
core IDs which Genode expects.
Ref:
https://developer.arm.com/documentation/100442/0100/register-descriptions/aarch64-system-registers/mpidr-el1--multiprocessor-affinity-register--el1?lang=en
* use netperf omnitest output selectors to acquire more metrics
* remove packet_size argument that was not interpreted by netperf
genodelabs/genode#4427
This does not affect default Genode builds as far as I can tell. There
is always at least one global static CTOR which seems to be coming from
one of the GCC runtime libs bundled in the toolchain. The problem became
visible for me only after I've replated GCC runtime with LLVM based
one. In such setup I often see binaries that do not have any static ctors.
Such binaries end up crashing Genode ld.lib.so.
Make sure the code does handle empty constructors array.
Fixes#4422
Switch the code to 0 lengh array instead. The code in
Lru_cache::element_size calls sizeof on this structure. This works in
gcc, but fails when using clang. Even for GCC however the documentation
states:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."
Basically its an implementation quirk that clang does not support. Both
GCC and clang do support zero sized arrays however so using them here
allows both compilers to process this code.
Ref: https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html
Issue #4421
This reference member is not used anywhere in the code. This prompts
clang to complain about it. Eliminate the member and all the plumbing
associated with it to silence the warning.
Issue #4421
Clang really doesn't like char subscripts. I can't say I blame it. Fix
the warning by an explicit cast to unsigned.
include/nitpicker_gfx/tff_font.h:230:53: error:
array subscript is of type 'char' [-Werror,-Wchar-subscripts]
Tff::Vertical_metrics const m = _vertical_metrics['m'];
^~~~
Issue #4421
Clang likes to complain when lambdas capture parameters without
actually using them. This patch fixes a couple of such problems in VFS
related os module code.
Issue #4421
The soft ABI implies purely software floating point implementation.
This is not the case for Genode however. For example core's
exception_vector.S uses vmsr instruction. This builds fine with with
GCC based toolchain, but clang with integrated-as complains:
src/core/spec/arm/exception_vector.S:122:2: error: instruction requires: VFP2
vmsr fpexc, r1
^
Fix this by passing softfp to mfloat-abi command on ARMv7. This allows
usage of FP HW, but implies soft-floating point ABI.
Issue #4421
According to C++11 reference:
"If the strictest (largest) alignas on a declaration is weaker than
the alignment it would have without any alignas specifiers (that is,
weaker than its natural alignment or weaker than alignas on another
declaration of the same object or type), the program is ill-formed:"
https://en.cppreference.com/w/cpp/language/alignas
The code requests 4 byte alignment for Genode::Arm_cpu::Context.
The Context structure inherits Genode::Arm_cpu::Fpu_context which
has minimum alignment requirement of 8 bytes, due to uint64_t d0_d31
member. This makes the 4 byte value in Context's alignas specifier
invalid (smaller than allowed minimum).
Similar situation takes place in Arm_64 case. The claimed minimum
alignment of Context is 8 bytes, but the fpu_state member imposes 16
bytes alignment (explicitly specified in Fpu_state declaration).
In both cases the code builds fine with GCC 8.3.0, but fails with
clang which claims that "requested alignment is less than minimum
alignment of X for type", where X is 8 on ARM and 16 on AArch64.
Ref: https://eel.is/c++draft/dcl.align#5
Issue #4421
This commit contains:
* Minimal Linux kernel target: pc_linux
* Library to generate a Linux build directory, config, generated headers
* API depot package
The actual work was provided by Josef Soentgen.
Ref genodelabs/genode#4416
The official way to obtain DMA addresses for RAM dataspaces is
the RPC function 'Pd_session::dma_addr' now. User-level device drivers
should not call this function directly but use the 'Platform_session'
interface of the platform driver instead.
Fixes#2243
* Creates sessions to all supported services of the black hole component
* Test-drives the Event and Capture session with dummy input
* Adds the test to the default list of depot_autopilot.run
* Test-driving the Audio_in and Audio_out sessions is still missing and should
be added via a dedicated commit
Ref #4419
To make room for the re-newed usb_host_drv basing on Linux 5.14 and
the re-newed lx_kit/lx_emul we have to move the depot recipe and
consistently name the old drivers with a legacy_ prefix.
Ref genodelabs/genode#4416
The x86 platform driver uses a different API than the one for ARM for
which the lx_kit glue code was designed. Since the x86 platform driver
will eventually adopt a similar interface we implement a wrapper that
encapsulates the old interface.
Ref genodelabs/genode#4411
Until now, the lx_emul layer addressed a 5.11 Linux Kernel port,
now that we add new architectures it is better to update the default version
first. There are especially changes in the task_struct code,
and the signature of some functions in the paging subsystem changed.
Ref genodelabs/genode#4411
We use the architecture-specific setjmp/longjmp implementation without
modification in the newer lx_kit implementation as well. There is no
need for a duplication.
Ref genodelabs/genode#4411
In the error case of socket_fs_accept() the Unconfirmed utility was
incompletely applied with the result of executing the cleanup routines
in the wrong order.
Fixes#4417
This patch removes the implicit build of ld-$(KERNEL) by the generic
ld.mk file because the kernel-specific dynamic linker is unreachable
when building a regular binary archive.
Issue #4320
When 'KERNEL' is specified, let the generic (pseudo) target of
lib/mk/ld.mk trigger the build of the actual dynamic linker named after
the used kernel. This way, we become able to remove the magic
linker-build step from the boot-image stage of the run tool.
Issue #4320
This patch applies the existing "privatization" of transitive
shared-library dependencies from static lib dependencies to shared lib
dependencies. It thereby improves the consistency of binaries created in
a regular build directory with binaries created in depot archives.
Issue #4408
This patch makes nitpicker's "clicked" report useful for the detection
of clicks outside of any client. This is needed in situations where the
dialog should close when clicking outside its screen area. In the new
version, a click outside any client results in a report without a
'label' attribute. Furthermore, the report is augmented by the sequence
number of the click, which allows for freshness checks by the consumer of
the report (i.e., sculpt_manager).
Issue #4398
If the buffer contains padding at the end, the iteration must continue
in order to restart iteration from the start of the buffer.
genodelabs/genode#4244
The archive contents are equivalent except the creation date of the
top-level directory.
-drwxr-xr-x guivol/users 0 2021-01-07 12:54 jpeg-9d/
+drwxr-xr-x guivol/users 0 2022-01-04 12:02 jpeg-9d/
Fixes#4406
The functionality of the test-block-client, test-block-server, and
test-block-bench components is now covered by the block_tester
application and the vfs_block server.
Issue #4405
This patch eliminates warnings that occurred as side effect of using the
'Session_policy' utility ("Warning: no policy defined for label...").
The new version uses the 'with_matching_policy' function instead, which
has the nice side effect of simplifying the error handling.
This patch makes the server-side policy-matching logic available outside
the 'Session_policy' class. Given that the new 'with_matching_policy'
function does not throw any exception, it gives server implementations
the freedom to avoid the C++ exception mechanism for the policy handling.
If the platform driver lacks the 'managing_system="yes"' attribute,
requests for DMA addresses return 0. This patch is meant to help
diagnosing such configuration issues.
Issue #2243
This patch enhances the PD-session interface with the support needed for
user-level device drivers performing DMA. Both RPC functions are
intended for the direct use by the platform driver only. If invoked for
PDs that lack the managing-system role, the operations have no effect.
The 'dma_addr()' RPC function allows the platform driver to request the
DMA address of a given RAM dataspace. It is meant to replace the
'Dataspace::phys_addr' RPC function.
The 'attach_dma' RPC function adds the given dataspace to the device
PD's I/O page table. It replaces the former heuristics of marking DMA
buffers as uncached RAM on x86.
With this patch, the UNCACHED attribute of RAM dataspaces is no longer
used to distinguish DMA buffers from regular RAM dataspaces.
Issue #2243
This patch makes Sculpt's leitzentrale GUI able to respond to touch events. It
formerly assumed that click/clack events are always preceded by hover reports
that identify the clicked-on widgets. For touch events, however, the most
up-to-date hover information referred to the previous click because there is no
motion without touching. So the GUI tended to identify the wrong widgets as
click targets.
The patch solved this problem by testing the freshness of the hover information
at the time of the click. If the hover information is older than the click, the
action is deferred until up-to-date hover information becomes available.
Fixes#4398
The new event type allows for the propagation of sequence numbers as a means to
validate the freshness of input handling. E.g., an menu-view-based application
can augment artificial sequence numbers to the stream of motion events supplied
to 'menu_view'. Menu view, in turn, can now report the latest received sequence
number in its hover reports, thereby enabling the application to robustly
correlate hover results with click positions.
Issue #4398
This patch replaces formerly blocking packet-stream operations by
the explicit use of 'wait_and_dispatch_io_signal' for blocking.
It also removes a misleading comment that promised a fire-and-forget
behavior whereas the implementation relied on blocking I/O anyway.
Issue #4390
Errors during IPC receive-and-wait can occur at the server side when
a client is killed. This condition is not an error from the server's
perspective. We used to print a message nevertheless, since the
condition is rather exceptional. However, when printed during the
test-sequence test, the messages interfere with the pattern matching of
the depot_autopilot, flagging the successful test as an error.
The VFS block plugin used to depend on the blocking semantics of the
packet stream's 'get_acked_packet'. This patch replaces this dependency
by the use of 'wait_and_dispatch_one_io_signal'. However, in order to
implement this change, the custom instance of a 'Signal_receiver' had to
be removed as well.
To keep this patch as little invasive as possible, it does not touch the
direct use of the block session's packet stream, which should better be
replaced by the 'Block::Connection::Job' API.
Issue #4390
The zynq_nic_drv follows a zero-copy approach and thus uses the packet
buffers as DMA memory. In order to know when the RX DMA memory can be used
for another packet, a custom ack_avail_handler is needed.
Similarly, packets received from the Uplink session are not copied to a
DMA buffer but to directly passed on as DMA memory. For this purpose,
a a custom packet_avail handler is needed.
genodelabs/genode#4384
The 'read' and 'write' utilities are from a time before the VFS API
as os/vfs.h was available. They rely on the (now removed) blocking
semantics of the packet-stream interface.
The only remaining legitimate use case of the direct interaction with
the file-system session without VFS is the back end of gcov, which needs
a way to exfiltrate the statistical data using a channel that is
independent from the libc or the VFS.
Issue #4390
This patch replaces the direct use of a file-system session via the
'file_system/util.h' helpers by the VFS using the os/vfs.h API. This
makes the component more flexible while removing the dependence from
read and write utilities of file_system/util.h, which happen to rely on
the (now removed) blocking packet-stream semantics.
Issue #4390
This patch replaces the direct interaction with the packet stream of
the block session by the use of the 'Block::Connection::Job' API,
removing the reliance on blocking packet-stream semantics.
Since I/O signals can now occur during 'Backend::submit', the patch
conditions the periodic calls of 'rump_sys_sync' by taking the backend
state into account.
Issue #4390
Since the change "block_tester: limit batching in sequential test", the
sequence test blocks infinitely when encountering a length value smaller
than the block size.
This patch takes precautions against the use of blocking packet-stream
operations like 'submit_packet'.
With the change of issue #4388, the ready-to-submit signals are no
longer implicitly handled. Hence, a call of submit_packet to a
saturated submit queue blocks infinitely.
Issue #4390
Even though the use of the C++ exception mechanism (and the implicit use
of the cxx heap) is not a problem at the server side, this patch
nevertheless replaces the exception-based return-value handling to make
the code consistent with the ipc_call path.
Issue #3612
* use 'Id_space::apply_any' instead of 'for_each' to destroy session local
buffer objects
* save session capability in session because 'cap()' is not valid in
'Session_component' destructor because it was dissolved before calling
'Root::_destruct_session', the cap is necessary to remove owned
buffers from the EP
issue #4380
Do not import a freshly allocated buffer to all contexts eagerly.
Instead check buffer list in context's 'exec_buffer' call and import
only buffers needed and not present before GPU execution. This leads to
improved performance for applications that use many OpenGL contexts
(e.g., VirtualBox 6).
issue #4380
* retrieve Genode::Env from plugin, this way no mesa applications need to
be changed.
* add 'vfs_gpu' api
* remove when all required functionality is implemented within the plugin.
issue #4380
Implement GLX and X functionality through Mesa's EGL interface. This
requires multiple OpenGL contexts and in turn GEM context support in
libdrm/iris, as implemented in libdrm and intel_gpu_drv.
Update recipes and machine.vbox6 files accordingly,
issue #4380
Retrieve multiple GPU sessions from VFS plugin, take advantage of buffer
import/export functionallity in order to implement gem context support.
Multiple contexts share all GPU buffers, but use different GPU sessions
and thus, differnt page tables and hardware contexts.
issue #4380
Add 'export_buffer'/'import_buffer' calls in order to support buffer
sharing between GPU sessions.
Reduce CAP costs for Gpu::Connection from 256 to 32.
issue #4380
Synchronize GPU completion by calling 'read' of the vfs_gpu plugin. This
enables pthreads to wait for GPU completions instead of the main EP.
issue #4380
When a <gpu> node is present in the VFS, the plugin opens a
Gpu::Connection for each open call to the 'gpu' node and installs a
completion signal handler. A read only to the fd returned only comes
back if a completion signal has been received between two read
transctions to the fd.
For now the Gpu::Connections can be retrieved by calling the
'vfs_gpu_connection(unsigned long id)' function. The id can be obtained
using 'stat' on the 'gpu' device and is located in the inode (st_ino)
field of the stat buffer.
issue #4380
* let iris handle buffer management, this implies that BOs are mapped to
the PPGTT during buffer execution and unmapped by iris later, for this
to work buffers need to be unmapped when allocating cached BOs
(vma_free) which requires a patch
* support lseek (drm_lseek for now) for determining object size
issue #4380
Account RAM and CAP resources for GPU sessions and trigger client
upgrades before allocating resources at the multiplexer. This prevents
the multiplexer from running out of resources.
issue #4380
Imagine receiving the signal for an available TX ack or an available RX packet
at the Uplink connection but a later received signal for a link-state change
(to link state "down") at the same connection is handled first and destructs
the Uplink connection before the handling of the former signals. In this case,
the methods 'Uplink_client_base::_conn_tx_handle_ack_avail' and
'Uplink_client_base::_conn_rx_handle_packet_avail' must be guarded against an
unconstructed '_conn' member, but they weren't so far.
Fixes#4384
So far, the generic Uplink connection code considered NIC drivers to transmit
connection RX packets in three different manners. Most of the drivers follow
the "normal" way of transmission with only one driver callback from generic
code. The monolithic USB NIC-driver, however, used to send in a "burst" mode
that required a dedicated path with multiple driver callbacks in the generic
code. And then there were drivers that had a fully custom function for doing
transmissions.
Also for handling connection TX acks, there was a "normal" and a
"custom handler" way.
Today, all NIC drivers in the Genode repos and the Genode-World repo follow the
"normal" way. Therefor, the unused code can be removed.
Ref #4384
Only parse output of the test programs one-by-one and ignore log
messages from other components not starting with '[init -> test$number]'.
Fixes sporadic failures on KERNEL=sel4 due to the following warning from
core.
Warning: flush page table entries - mapping cache full - PD: init -> test1
To actually receive audio input in a particular VM, it also
needs to be enabled in the vbox file:
<AudioAdapter controller="HDA" driver="OSS" enabled="true"
enabledOut="true" enabledIn="true"/>
Fixes#4377
On some boards or emulators a CPU might be executing in hyp mode
when entering Genode's bootstrap code. In that mode the 'cps' instruction
is not defined. Therefore, we change the way the boot cpu is identified.
Ref #3415
* renamed rpi pic to Bcm2835_pic
* renamed rpi3 pic to Bcm2837_pic
* added bcm2837 control for setting prescaler value (to fix timer_accuracy)
* changed handling of all interrupts for rpi3 by cascading to bcm2835 pic
* rpi3 irq controller base address made consistent with rpi
* added usb controller memory region for pic on rpi3 (for SOF interrupts)
Ref #3415
The socket type (in the lower bits) maybe ORed with SOCK_CLOEXEC and
SOCK_NONBLOCK options (in the higher bits). Currently, supported values
are SOCK_STREAM (1) and SOCK_DGRAM (2), so just take the lower 2 bits.
This fixes treating `SOCK_STREAM` sockets as UDP if additional flags
were set.
Fixes#4370
This patch equips Sculpt with the ability to customize the system image
in very flexible ways.
All customizable aspects of the image have been relocated from the
former sculpt.run script and the accompanied gems/run/sculpt/ directory
to a new location - the sculpt/ directory - which can exist in any
repository. The directory at repos/gems/sculpt/ serves as reference.
The sculpt directory can host any number of <name>-<board>.sculpt files,
each containing a list of ingredients to be incorporated into the
Sculpt system image. The <name> can be specified to the sculpt.run
script. E.g., the following command refers to the 'default-pc.sculpt'
file:
make run/sculpt KERNEL=nova BOARD=pc SCULPT=default
If no 'SCULPT' argument is supplied, the value 'default' is used.
A .sculpt file refers to a selection of files found at various
subdirectries named after their respective purpose. In particular, There
exists a subdirectory for each file in Sculpt's config fs, like
nitpicker, drivers... The .sculpt file selects the alternative to use
by a simple tag-value notation.
drivers: pc
The supported tags are as follows.
*Optional* selection of /config files. If not specified, those files are
omitted, which prompts Sculpt to manage those configurations
automatically or via the Leitzentrale GUI:
fonts
nic_router
event_filter
wifi
runtime
gpu_drv
Selection of mandatory /config files. If not specified, the respective
'default' alternative will be used.
nitpicker
deploy
fb_drv
clipboard
drivers
numlock_remap
leitzentrale
usb
system
ram_fs
Furthermore, the .sculpt file supports the optional selection of
supplemental content such as a set of launchers.
launches: nano3d system_shell
Another type of content are the set of blessed pubkey/download files
used for installing and verifying software on target.
With the new version, it has become possible to supply a depot with the
the system image. The depot content is assembled according to the 'pkg'
attributes found in launcher files and the selected deploy config.
The resulting depot is incorporated into the system image as 'depot.tar'
archive. It can be supplied to the Sculpt system by mounting it into the
ram fs as done by the 'ram_fs/depot' configuration for the ram fs.
It is possible to add additional boot modules to the system image. There
are two options.
build: <list of targets>
This tag prompts the sculpt.run script to build the specified targets
directly using the Genode build system and add the created artifacts
into the system image as boot modules.
import: <list of depot src or pkg archives>
This tag instructs Sculpt to supply the specifid depot-archive content
as boot modules to the system image. This change eliminates the need for
board-specific pkg/sculpt-<board> archives. The board-specific
specializations can now be placed directly into the respective .sculpt
files by using 'import:'.
To make the use of Sculpt as testbed during development more convenient,
the log output of the drivers, leitzentrale, and runtime subsystems
can be redirected to core using the optional 'LOG=core' argument, e.g.,
make run/sculpt KERNEL=linux BOARD=linux LOG=core
The former pkg/sculpt-installation and pkg/sculpt-installation-pc
archives have been replaced by pkg/sculpt_distribution-pc, which
references the generic pkg/sculpt_distribution archive. Those pkgs are
solely used for publishing / distribution purposes.
Fixes#4369
When used by the 'Allocator_avl' the slab allocator's backing store is
dynamically disabled and re-enabled while adding/freeing ranges.
However, during those operations, slab entries can be freed. This,
in turn, can result in the release of a slab block (when the freed slab
entry happens to be the last entry of the block). In this corner case,
'Slab::_release_backing_store' operation has no effect because no
backing-store allocator is set. As a result, the block is no longer
referenced but not physically freed.
The patch fixes the problem by skipping '_free_curr_sb' whenever
no backing store is defined. So the completely empty block remains
in the working set.
Thanks to Peter for reporting and fixing this issue!
Fixes#4367
- Consider 'sdl' as source of input events in the event-filter
configuration as generated by the sculpt manager
- Supply an artificial 'platform_info' ROM as requested by the
sculpt manager to obtain the affinity-space information
- Substitute 'fs_rom' for 'cached_fs_rom' as a workaround for the
lack of support for managed dataspaces on Linux
Fixes#4362
This change keeps the version-controlled 'pubkey' and 'download' files
separate from files generated via depot/create or downloaded via
depot/download. So one can remove the entire depot/ directory without
interfering with git.
Furthermore, depot keys can now be hosted in supplemental repositories
independent from Genode's main repository.
Fixes#4364
Fix some trivial cases where the signedness of the constant value does
not match the signedness of type the code expects to see. GCC can be
asked to warn about those by passing Wsign-covnersion flag.
Issue #4354
This comes up when building the code with clang 13. It happens due to
recently enabled Wconversion warning, which in case of clang also
enables implicit-int-conversion warning. The warning reads:
fs_file_system.h:937:44: error: higher order bits are zeroes after
implicit conversion [-Werror,-Wimplicit-int-conversion]
::File_system::Watch_handle fs_handle { -1U };
~~~~~~~~~ ^~~
This can be fixed by properly specifying fs_handle value to be of
unsigned long type.
Issue #4354
As far as I can tell this is not raised by any released GCC versions.
Clang 13 on the other hand warns about it due to implicit-int-conversion
warning which is automatically enabled together with Wconversion. The
problem is relatively simple, shifting access_t value does not always
produce result which is also of access_t type. For example, if access_t
is uint16_t, shifting it will produce integer result. This can be
observed even with GCC. Building the following C++ example will fail:
#include <type_traits>
#include <stdint.h>
int test() {
uint16_t a = 0xabcd;
static_assert(std::is_same_v<decltype(a<<1), uint16_t>);
return 0;
}
Changing uint16_t in the static_assert to int, will allow the code to
build.
Make such int to access_t implicit conversion explicit to allow the code
to be compiled with both GCC and clang.
Issue #4354
The path for the `nameserver` file was fixed to `/socket/nameserver`.
So, if the socket directory was configured differing from `/socket`, DNS
did not work. Now the default path for the `nameserver` takes into
account the path configured in the `socket` libc config attribute.
Fixes#4318Fixes#4343
Remove '_expected_offset' check on round trip test from RX packet,
because allocators between RX/TX can have different allocation
strategies. Rely on 'pattern' check for RX packets only.
In loopback server alloc size must match actual packet size.
issue #4312
Override 'try_alloc/free' because ethernet frame headers are 14 bytes
(src/dst mac (12) + ethertype (2)) causing the IP header to be 2 byte
aligned, leading to problems on platforms that require load/store
operations to be naturally aligned when reading, for example, 4 byte IP
addresses. Therefore, we align the allocation to 2 bytes, so the IP
header is aligned to 4.
issue #4312
This patch improves the robustness of the CPU-affinity handling.
- The types in base/affinity.h received the accessors
'Location::within(space)' and 'Affinity::valid', which alleviates
the fiddling with coordinates when sanity checking the values,
in init or core.
- The 'Affinity::Location::valid' method got removed because its
meaning was too vague. For sanity checks of affinity configurations,
the new 'within' method is approriate. In cases where only the x,y
values are used for selecting a physical CPU (during thread creation),
the validity check (width*height > 0) was not meaningful anyway.
- The 'Affinity::Location::from_xml' requires a 'Affinity::Space'
as argument because a location always relates to the bounds of
a specific space. This function now implements the selection of
whole rows or columns, which has previously a feature of the
sandbox library only.
- Whenever the sandbox library (init) encounters an invalid affinity
configuration, it prints a warning message as a diagnostic aid.
- A new 'Affinity::unrestricted' function constructs an affinity that
covers the whole affinity space. The named functions clarifies
the meaning over the previous use of the default constructor.
- Core's CPU service denies session requests with an invalid
affinity parameter. Previously, it would fall back to an
unrestricted affinity.
Issue #4300
Instead of using a bitfield for storing rwx and skip boolean value,
take a boolean instead. This fixes a note giv]en by GCC 9.1 about
changes semantics of bitfields given as parameter by value on ARM.
Ref #4344
Similar to other devices allow for selecting a graphics device as
well. This is currently only useful on iMX8-based systems like
iMX8Q-EVK and MNT Reform2 where the driver is explicitly started
from a pkg.
Fixes#4342.
After VirtIO::Queue refactoring buffers no longer share the same
dataspace as VirtIO rings. This makes optimal buffer calculations a lot
easier. In this case 64 buffers 2kB each will need precisely 128kB of
RAM. Previous value of 2016 will just waste 768b.
Fixed#4347
The key changes in this patch are:
* Buffer allocation is moved into a separate Buffer_pool helper. The
implementation of the buffer allocation strategy does not change.
The helper allocates a single RAM dataspace and splits it in multiple,
equally sized chunks.
* Management of main descriptor ring is enacapsulated in Descriptor_ring
helper class.
* Use separate RAM dataspaces for descriptor rings and buffers.
Previously both of them were packed into a single dataspace. This
might have been more RAM efficient, but IMO it made the code uglier and
harder to understand.
* All of the VirtIO::Queue members are now initialized on the class member
initializer list. This is possible due to previously listed changes.
* Since all VirtIO::Queue members are initalized on member initalizer
list, some additional ones can be marked as const, ex _avail, _used ring
pointers.
* Move descriptor writing code into a common method used by both
write_data and write_data_read_reply members. This avoids some code
duplication between those methods.
* Get rid of request_irq argument that most public VirtIO::Queue methods
accept. None of the existing drivers use it and I doubt this will
change any time soon.
* Use Genode namespace by default in Virtio.
This patch also fixes at least one bug that I discovered while working
on VirtIO block device driver. Namely, when chaining descriptors only the
first descriptor in the chain should be exposed in the available ring.
Issue #4347
The const-variant of the data() method contained an erroneous
calculation of the tail size. This led to the size guard throwing
exceptions when trying to parse TCP packets that only contained the
TCP header.
Fixesgenodelabs/genode#4340
Thanks to Piotr Tworek for the fix and his explanation as follows:
The basic idea is to try to fit payload data into the descriptor used to
send the header. If there is no payload, or the payload fits exactly
into the remaining space in the header decriptor, len should be 0 and
only one descriptor should be used. In such case the "next" and "flags"
members of the descriptor structure should be set to 0.
In case there is some extra payload data to send, but its size is
bigger than the remaining free space in the descriptor used to send the
header, len should contain the remaining size of the payload that
can't be sent via the header descriptor. The code will then chain
additional descriptors to handle this remainder.
With the len variable shadowing, the code will never queue the remaining
data.
Issue #4327
Share datastructures for clock, power and reset related configurations
per device. In the generic platform driver component these structures
are kept empty. Driver derivates can fill the clocks settings, power and
reset switches with life. The former Driver::Env gets removed.
Fix#4338
This change of the inner working of the platform driver for ARM allows
clients to have permanent open sessions, as long as a policy node matches
the client. If devices disappear from the policy resp. from the set of
available devices (hotplug), the devices ROM of the session gets updated,
and a corresponding device session gets closed. If the device remains
untouched in the configuration but other devices appeared/disappeared, the
device session is not affected.
Ref #4330
The requested guest-physical memory range may comprise multiple attached
dataspace regions, which must all be detached. This is not required for
the current vbox5 implementation, but for vbox6 as the current API
suggests these semantics.
This commit can be seen as intermediate fix as a real fix should change
the API to prevent long-running detach loops in core that may lock out
requests by other components.
* Only give managing_system permission when all parent nodes of the
corresponding component agree in doing so.
* Move the physical memory constrains heuristic from sandbox library to core
Fix#4335
* Track all caps and ram quotas of the sub-sessions properly
* Release DMA buffers, it is not done implicitely when destroying
the Constrained_ram_allocator
* Do not replenish quota before really releasing memory from
the allocator
Issue #4330
The Session_component must be destroyed before updating the device
model because the Session_component must also release all previously
acquired devices. If the device model is updated before, the devices
might have been removed.
Issue #4330
Pre-allocate all possible type of policy objects as part of the thread meta
state to avoid increased memory consumption due to different policy object
sizes. The cpu_balancer accounts the memory per client and can't forward
potentially occurring out-of-ram exceptions during config-ROM update phases.
Fixes#4333
The commits avoids reading in and allocating memory for all potentially
threads, which are potentially currently not existent (but configured in the
policy beforehand). Instead the policy is read in and evaluated when a thread
is created and policy changes are solely applied to existing/running threads.
By this the commit avoids the increase of memory consumption during the
evaluation of policies during config ROM updates.
Issue #4333
This implements the necessary bits to provide 2D framebuffer support on
top of VirtIO GPU device as implemented in Qemu. I don't know if any
other implementation of this specific device exists.
Compared to the ramfb driver which already exists in Genode Virtio FB driver
has one major benefit. It allows Qemu window to be dynamically resized at
runtime. The driver will treat this as resolution change and act accordingly.
Ramfb driver can currently only use the hardcoded 1024x768 screen size. Changing
screen resolution might not sound like a big deal, but it is rather useful to
run Genode on Qemu in full screen mode.
Some more advanced devices like VirtIO GPU do expect they can receive
responses to VirtIO commands they issue via VirtIO queue. Such responses
are not sent via a separate device writeable queue. Instead the driver
is expected to queue some additional descriptors and buffers which the
device can then use to provide the reply.
This patch adds support for such write-data-read-response opeartion to
Genode VirtIO::Queue implementation. The implementation is pretty simple
and does not support any fancy features like receiving the response
asynchronously. Instead the operation will use caller provided callback
to wait for the device to process the command. Once this callback
returns the write-data-read-response VirtIO::Queue function will invoke
another callback passing received response as argument.
Mesa queries information about the underlying device and this header
denotes the layout of the information. It is also used by the driver
itself to populate the 'info_dataspace'.
Issue #4329.
Since the parts of the 'etnaviv' library are already part of the
'mesa.lib.so' that is normally loaded along-side remove the
duplication here.
Issue #4329.
This filter bridges the gap between a touchscreen driver, which
generates raw touch events and traditional GUI applications that expect
a pointer (absolute motion, press/release of the left mouse button).
Fixes#4332
This patch changes the 'Allocator' interface to the use of 'Attempt'
return values instead of using exceptions for propagating errors.
To largely uphold compatibility with components using the original
exception-based interface - in particluar use cases where an 'Allocator'
is passed to the 'new' operator - the traditional 'alloc' is still
supported. But it existes merely as a wrapper around the new
'try_alloc'.
Issue #4324
The former use of Pthread conditionals did not cover the corner case of
early wakeups just before halting the CPU. These wakeups were simply
lost which resulted in sporadic halts of about 500 ms (the maximum timeout
of all halts in VirtualBox). RTSEMEVENTMULTI preserves early wakeups
and effectively prevents the CPU from halting.
Additionally, we now wakeup the target CPU on VMMR0_DO_GVMM_SCHED_POLL
and, thus, mimic the behavior of the original implementation slightly
better,
Slightly related to #4313
Since the top-level node of the output ROM is always generated by the
rom_filter, there is no way to pass-through the content of an input ROM
without wrapping in an addition XML node.
genodelabs/genode#4326
The Allocator_avl back end will display diagnostic messages if the
address to be freed is not at the beginning of a block. This happens
regulary when 'struct page' objects are not freed in allocation order.
Issue #4325.
Introduce a method to access the dataspace capability of the underlying
backing store for a memory allocation. This is necessary for drivers
where the memory is managed manually and the capability needs to be
given to a client.
Issue #4325.
Explicitly, adapt to current framebuffer/window size after
initialization finished. This ensures the use of the correct framebuffer
dimensions in scenarios without a window manager.
Thanks to Raphael for the patch.
This patch makes the test less dependent on the rate of state updates by
calculating the upgraded quota from the values found in the state report
instead of simply increasing the '_ram_quota' for each incoming report.
This patch replaces the 'Ram_allocator::alloc' RPC function by a
'try_alloc' function, which reflects errors as 'Attempt' return value
instead of an exception.
Issue #4322
Issue #3612
The new 'update_list_model_from_xml' function template simplifies the
use of the list model utility by alleviating the need for implementing a
custom policy class for each model. Instead, the transformation is done
using a few lambda functions given directly as arguments.
Issue #4317
This patch introduces the lx_emul/pin.h interface that enables GPIO stub
drivers to interact with Genode's Pin_control and IRQ sessions via a
simple C API.
Fixes#4316
The new interfaces are meant to gradually replace the existing
Gpio_session interface.
- Each session refers to a single pin.
- The session types distiguish the direction of the signal as input or
output.
- Pin coordinates can be selected via session labels.
- GPIO interrupts are covered by the regular IRQ session interface.
The interfaces are accompanied by framework utilities and interfaces:
- os/pin_driver.h
- pin_control_session/component.h
- pin_state_session/component.h
These headers relieve GPIO drivers from implementing boilerplate code by
providing device-agnostic portions. The A64 pio driver serves as
reference for using those utilities.
https://github.com/nfeske/genode-allwinner/tree/master/src/drivers/pin/a64Fixes#4315
Always instantiating a network device with id `net0`, removes the need to call
append_qemu_nic_args in run scripts unless we want to add forwarding
rules.
genodelabs/genode#4311
Allow specifying additional qemu arguments for externally supported boards
(e.g. zynq_qemu) by adding a `qemu_args` file in the board-property directory.
The syntax of the qemu_args file is as follows:
- Arguments can appear in a single line or in multiple lines as the
lines will be appended (separated by a whitespace) to the global
qemu_args variable.
- If the line is prepended with a `foobar:` expression. The arguments
are only added if the foobar spec is present.
Note, that a `-m` argument specified in the qemu_args file will
override the arguments provided by the run scripts.
genodelabs/genode#4311
First, the former implementation has only considered the pure numerical
variant of the -m argument. Yet, qemu also allows specifying the amount
of memory by `-m 1G`, `-m size=1G` and more.
Second, the default amount of memory for BOARD=pc was 512M (800M in case of okl4).
Since the depot_autopilot.run also required at least 768M on all
platforms it seems reasonable to take 800M as a default value for BOARD=pc and
thereby remove the special treatment of okl4.
genodelabs/genode#4311
- request FPU state on VM exit in portal config
- transfer FPU state on VM entry
- save fpu state early
Avoid any FPU instructions (for example during base API calls), which
use the FPU and overwrite FPU registers holding the guest vCPU FPU
state.
- don't save fpu state of EP
Issue #4313
The commit avoids the race between checking for the NEM state and the decision
to notify (poke) the NEM backend (nem.cc).
- ever notify about POKE flag for remote vCPU
- check synchronized for NEM state
- store recheck flag and apply on next switch_to_hw
Issue #4313
In the default setting the number of root ports is set to 4. This
also limits the number of USB host devices that may be attached to
the VM. Since the USB webcam shares the slots, that leaves us with
only 3 available USB2 slots to attach USB host devices. Depending on
the use-case that might not be enough.
This commit statically increases the number of ports to 8 each and
adapts the qemu-usb glue-code accordingly.
Many thanks to Raphael for initial investigation and workaround.
Fixes#4310.
Accidentally, the size of IOMEM dataspace got accounted within the
implementation of the platform driver for ARM. Instead we should
only account a bit for the metadata and paging of the I/O memory.
Fix#4307
The custom ack avail handler is required for zero-copy nic drivers (e.g.
the zynq nic driver), which must release the corresponding DMA buffers.
Fixesgenodelabs/genode#4277
This patch adds a missing call of '_handle_decorator_margins' at
construction time. Up to commit "report_rom: versioning and explicit
notification" this problem was masked by an unconditional signal, which
implicitely triggered the call.
Issue #4274Fixes#4306
This component can service Qemu VirtIO mouse, keyboard and tablet
devices. The implementation is based on VirtIO 1.1 device spec. Its
described in section 5.8 "Input Device".
Issue #4282
Those symbols are not satisfied by any code or dependency of libm. As
result calling cprojf function will always crash on Genode. This crash
can be turned into link time error by adding --no-undefined to LD_OPT.
This patch provides the missing symbols by including isninf.c in libm
build.
Fixes#4299
When rebasing my local branch on top of sculpt-21.10 tag I've noticed
two problems.
The code in new_delete.cc does not include new header file. This works
fine with GCC, but fails with clang because std::align_val_t type is
not defined anywhere according to clang. It looks like GCC pulls this
header indirectly somehow.
The second problem can be seen if one disallows undefined symbols in
executables and shared_libraries. This can be seen with both GCC and
clang by adding --no-undefined to LD_OPT. With such change in place core
fails to link due to:
ld.lld: error: undefined symbol: operator delete(void*, std::align_val_t)
>>> referenced by thread.h:448 (/home/tworaz/devel/genode/repos/base-hw/src/core/kernel/thread.h:448)
>>> thread.o:(Kernel::Core_main_thread::~Core_main_thread()) in archive debug/core-hw-virt_qemu.a
>>> referenced by thread.h:448 (/home/tworaz/devel/genode/repos/base-hw/src/core/kernel/thread.h:448)
>>> thread.o:(non-virtual thunk to Kernel::Core_main_thread::~Core_main_thread()) in archive debug/core-hw-virt_qemu.a
>>> did you mean: operator delete(void*, unsigned long, std::align_val_t)
>>> defined in: debug/core-hw-virt_qemu.a(supc++.o)
If the code would somehow manage call such undefined symbol it'd crash.
Since I generally prefer link time failures to runtime crashes I link
all genode binaries with --no-undefined.
To fix this problem just add a dummy implementation of missing delete
operator.
Fixes#4298
Alignas should be placed before the type. Placing it after it works for
GCC, but fails when building the same codee with clang. The error
message is:
reconstructible.h:48:27: error: 'alignas' attribute cannot be applied to types
char _space[sizeof(MT)] alignas(sizeof(addr_t));
^
Issue #4298
Apparently the iris driver does not make use of tiling by the kernel, so
we shortcut the 'SET_TILING' call to keep iris happy with this quickfix.
However, tiling information may get lost, if the iris driver ever calls
'MMAP_GTT' and no fence is configured for the buffer. A follow-up commit
should address this shortcoming in the future.
Issue #4284
- CPU-architecture annotations
- Change order of top-level menu, moving Mesa driver to the end
- Add black_hole, recall_fs, file_fault
- Add usb_webcam, test-capture
- Add audio driver and mixer
- Add vbox6, keeping vbox5-nova-sculpt as fallback
- Remove recall_fs launcher, which is obsolete with the recall_fs pkg
- Replace system_clock by system_clock-pc pkg
Issue #4281
This follow-up commit to "sculpt: avoid flickering of leitzentrale"
allows nitpicker to double-buffer pixels during resize operation on a
screen size of 1920x1200.
This patch presents all press and release events to the pointer state,
fixing the problem that _key_cnt was decreased but never increased.
However, the inconsistency had no observable effects in practice.
Issue #4176
This patch extends the notion of having only one uniquely hovered client
in the presence of held keys.
If motion occurs once a key is pressed (e.g., while dragging), the
receiver of the key sequence observes the motion events. In this case,
we have to submit an artificial leave event to the originally hovered
client so that no more than one client observes itself as being hovered
at the same time. Once the key sequence is finished, the hovering is
updated again, eventually presenting a motion event to the originally
hovered client and a leave event to the receiver of the key sequence.
Issue #4176
This patch ensures that the pointer report is updated not before all
input events are handled. The change does not solve any observed
practical issue but the potential problem was spotted while reviewing
the code.
Issue #4176
To clearly identify the correct device use the unique device's name
not the type. Otherwise a driver cannot drive several devices of the
same type.
Fix#4297
This commit contains a backport of commit [1] that deals with updating
the event ring dequeue pointer more often to prevent unnecessary
'Event Ring Full' errors.
[1] 'usb: host: xhci: update event ring dequeue pointer on purpose'
(dc0ffbea5729a3abafa577ebfce87f18b79e294b)
Fixes#4296.
This patch changes the depot_query tool to filter the returned index
data depending on the 'arch' as specified for the query. This way, one
index file can support multiple CPU architectures while allowing
individual entries to be architecture-specific.
Fixes#4295
This patch adds the missing definition of 'prio_levels' in the prepare
sub init, fixing the warning:
[init -> runtime -> prepare] Warning: vfs: invalid priority, upgrading from -2 to 0
Issue #4281
Use 'StateChange' event to check for machine's 'PowerOff' state, close
Gui connections and submit exit signal to EP which in turns calls exit.
Fixes#4291
The new 'Env::try_session' method mirrors the existing 'Env::session'
without implicitly handling exceptions of the types 'Out_of_ram',
'Out_of_caps', 'Insufficient_ram_quota', and 'Insufficient_cap_quota'.
It enables runtime environments like init to reflect those exceptions to
their children instead of paying the costs of implicit session-quota
upgrades out of the own pocket.
By changing the 'Parent_service' to use 'try_session', this patch fixes
a resource-exhaustion problem of init in Sculpt OS that occurred when
the GPU multiplexer created a large batch of IO_MEM sessions, with each
session requiring a second attempt with the session quota upgraded by
4 KiB.
Issue #3767
This commit removes the report service from the usb_webcam pkg, which
was used to enable or disable the webcam driver. The on/off state is
instead controlled by the presence of a capture client. That is, once a
capture client appears, the webcam driver is started. Vice versa, once
no capture client exists, the webcam driver is removed automatically.
Internally, the detection of presence of a capture client is based on
nitpicker's 'displays' report, which is consumed as input of the
rom_filter, which in turn generates the configuration of a dynamic sub
init.
Fixes#4287
There are some subtle incompatibilities in VirtualBox 6 with settings we
used in version 5. Therefore, the vbox6 package uses machine.vbox6 as
configuration file. An example configuration is provided by the
raw/vbox6 package.
Issue #4281
'Out_of_ram' was so far the only exception a client had to deal with
during buffer managment. Allocating memory, however, does not only
consume RAM quota but CAP quota as well.
This commit tries to mitigate that shortcoming by reflecting the
'Out_of_caps' state back to the client. Furthermore it allows for
resource accounting on certain client allocations, e.g. buffers.
Fixes#4284.
Allocating and mapping buffers not only consumes RAM quota, it consumes
CAP quota as well. Extended the Gpu session to allow for dealing with
that on the client side.
On a side note, the amount of initial CAP quota needed to establish
a connection is increased to cover the current costs of the Intel
GPU multiplexer.
Issue #4284.
The platform driver uses a 'Constrained_ram_allocator' to allocate
meta-data on behave of a client. It uses the PD session as
'Ram_allocator' back end that in return is implemented via the
'Expanding_pd_session_client'.
Whenever the PD client itselft comes into resource shortage it will
ask its parent unconditionally. However, depending on the integration,
such a request may be left unanswered.
This commit introduces a check to prevent that situation from
occurring. In case the platform driver notices the resource shortage
it will reflect that back to the client.
Issue #4284.
By using the cached_fs_rom instead of fs_rom, each client obtains the
same (read-only) copy of the shared library, avoiding a RAM resource
request when running mulitple GPU applications at the same time.
With this patch, Sculpt is able to host at least 3 Mesa applications.
Issue #4263
Issue #4281
This new version of the system_clock pkg does no longer depend on the
presence of an external 'Rtc' service as previously provided by the
Sculpt base system. Instead, it hosts the rtc_drv inside the subsystem.
Because rtc_drv is board-dependent, the system_clock pkg is named
system_clock-pc now.
Issue #4281
Some guests don't handle remote wake up correctly causing devices to
stop functioning. Therefore, we disable the remote wake up bit (5) in
`bmAttributes` of the device configuration descriptor.
Thanks to Peter for the initial fix.
Fixes#4278
ROM clients have to request an initial update of dynamic ROMs explicitly
and should not depend on artificial signals from the ROM session on
signal-handler registration.
Issue #4274
The sequence app should immediately stop the child when it called
parent().exit(). Otherwise, the child will continue execution which
causes a race condition: The child's ld.lib.so will eventually destruct
an Attached_rom_dataspace for the config rom. If sequence destructed the
corresponding service first, we will get an Ipc_error.
genodelabs/genode#4267
Warning!
The current version of the file vault is not thought for productive use but
for mere demonstrational purpose! Please refrain from storing sensitive data
with it!
The File Vault component implements a graphical frontend for setting up and
controlling encrypted virtual file systems using the Consistent Block Encrypter
(CBE) for encryption and snapshot management. For more details see
'repos/gems/src/app/file_vault/README'.
Fixes#4032
Previously unconditional calls to Genode::log in cbe init and the cbe trust
anchor VFS plugin were made dependent on a verbosity flag that is set to
"false" by default.
Ref #4032
Instead of simply encrypting the private key with AES-256 when storing it to
the 'encrypted_private_key' file, wrap it using the AES-key-wrap algorithm
described in RFC 3394 "Advanced Encryption Standard (AES) Key Wrap Algorithm".
This is more secure and enables us to directly check whether the passphrase
entered by the user was correct or not.
Ref #4032
As the file formerly named 'secured_superblock' actually contains the hash of
the superblock that was secured, it was renamed 'superblock_hash'.
Ref #4032
As the file formerly named 'keyfile' actually contains the encrypted private
key of the Trust Anchor, it was renamed 'encrypted_private_key'.
Ref #4032
By now, the symmetric keys were only XOR'ed with the private key as placeholder
for a real encryption. Now they are encrypted using AES256 with the TA's
private key as key
Ref #4032.
A private key of 256 bits is generated pseudo-randomly using the jitterentropy
VFS plugin on initialization. The private key is stored in the key file
encrypted via AES256 using the SHA256 hash of the users passphrase. When
unlocking the CBE device, the encrypted private key is read from the key file
and decrypted with the hash of the users passphrase.
Ref #4032
Instead of using the user passphrase directly, use its SHA256 hash calculated
using libcrypto. The passphrase hash is still stored in the key file to be
used as base for the very primitive way of generating the private key.
Ref #4032
Closing the keyfile handle after a write operation wasn't synchronised to the
actual end of the write operation.
Issuing a write operation at the back end returns successfull as soon as the
back end has acknowledged that it will execute the operation. However, the
actual writing of the data might still be in progress at this point. But the
plugin used to close the file handle and declare the operation finished at this
point which led to warnings about acks on unknown file handles and leaking
resources. Now, the plugin issues a sync operation directly after the write
operation and waits for the sync to complete. This ensures that the plugin
doesn't declare the operation finished too early.
Ref #4032
The unlocking operation in the trust anchor was broken wich caused bad keys in
the CBE. This rewrites the whole operation to work as desired. Note that this
doesn't make it more safe! The private key is still almost the same as the
passphrase and stored plaintext.
Ref #4032
The plugin used to close file handles via the 'vfs_env.root_dir.close'.
However, this lead to resource leaks and apparently isn't the right way to
do it. Other VFS plugins do it by calling 'close' directly on the handle and
doing it in the trust anchor plugin also, fixes the leaks.
Ref #4032
Closing the hashfile handle after a write operation wasn't synchronised to the
actual end of the write operation.
Issuing a write operation at the back end returns successfull as soon as the
back end has acknowledged that it will execute the operation. However, the
actual writing of the data might still be in progress at this point. But the
plugin used to close the file handle and declare the operation finished at this
point which led to warnings about acks on unknown file handles and leaking
resources. Now, the plugin issues a sync operation directly after the write
operation and waits for the sync to complete. This ensures that the plugin
doesn't declare the operation finished too early.
Ref #4032
There were no means for issuing a Deinitialize request at the CBE using the
CBE VFS plugin. The new control/deinitialize file fixes this. When writing
"true" to the file, a Deinitialize request is submitted at the CBE. When
reading the file, the state of the operation is returned as a string of the
format "[current_state] last-result: [last_result]" where [current_state] can
be "idle" or "in-progress" and [last_result] can be "none", "success", or
"failed".
Ref #4032
When discarding a snapshot, the CBE VFS plugin didn't communicate the ID of
the snapshot to the CBE. Instead it set the ID argument to 0. Therefore the
operation never had any effect.
Ref #4032
The snapshots file system used to return the number of snapshots on
'num_dirent' when called for the root directory although it was expected to
return 1. This confused the tooling ontop of the VFS.
Ref #4032
Despite being readable, the files control/extend and control/rekey proclaimed
that they were not when asked. This caused the fs_query tool to not report the
content of the files although it could have.
Ref #4032
Stat calls on the control/extend and control/rekey files returned a bogus file
size that led to an error in the VFS File_content tool. The tool complained
that the size of the file determined while reading the content differs from the
one reported by the stat operation. Now, the stat call will always determine
the actual size of what would be read. However, it isn't guaranteed that this
size doesn't change in the time after the stat operation and before the read
operation.
Ref #4032
The service is loaded dynamically VBoxSharedClipboard.so at runtime. The
VFS configuration mounts the shared object at /VBoxSharedClipboard.so as
the file is checked by contrib code before loading. An init
configuration in pkg/vbox6/runtime illustrates this and how to re-label
the VBoxSharedClipboard.so ROM to its real name
virtualbox6-sharedclipboard.lib.so.
During Windows 10 boot with sequential block requests, the AHCI request
worker finished earlier than the EMT thread signals hEvtProcess and
begins waiting for hEvtProcessAck indefinitely. The timeouts helps to
survive this short phase.
A better solution would use conditional variables, which are not
provided in VirtualBox's runtime.
This patch introduces a C API to be used by input drivers to generate
Genode events. The initial version is limited to multitouch events only.
Fixes#4273
* Use the architecture-dependent minimal alignment for all allocations,
e.g. on ARM it is necessary to have cacheline aligned allocations for DMA
* Remove the allocation functions without alignment from generic API
* Fix a warning
Fix#4268
After a DMA transaction do only invalidate cachelines from the
corresponding DMA buffers if data got transfered from device to
CPU, and not vice versa. Otherwise it might result in data corruption.
Ref #4268
The former implementation did not internally track ROM changes notified
vs. delivered to the client. We adapt the versioning implementation
implemented in dynamic_rom_session.h and enable explicit notification of
the current version.
The feature is used by the clipboard to notify permitted readers of the
clipboard ROM service on focus change via the newly created private
Rom::Module::_notify_permitted_readers() function.
Fixes#4274
The includes for the address-space-ID allocator and the translation table are
usually specific to the CPU in use. Therefore these includes can be moved from
their current location in the board header to the CPU headers. This reduces the
number of decisions a board maintainer has to make if the CPU model he's aiming
for is already available.
This can probably also be applied for other includes in the board headers but I
intentionally leave it for a future commit as I don't have the time to do it
all now.
Ref #4217
For base-hw Core, we used to add quite some hardware-specific include paths
to 'INC_DIR'. Generic code used to include, for instance, '<cpu.h>' and
'<translation_table.h>' using these implicit path resolutions. This commit
removes hardware-specific include paths except for
1) the '<board.h>' include paths (e.g., 'src/core/board/pbxa9'),
2) most architecture-specific include paths (e.g., 'src/core/spec/arm_v7'),
3) include paths that reflect usage of virtualization or ARM Trustzone
(e.g., 'src/core/spec/arm/virtualization').
The first category is kept because, in contrast to the former "spec"-mechanism,
the board variable used for this type of resolution is not deprecated and the
board headers are meant to be the front end of hardware-specific headers
towards generic code which is why they must be available generically via
'<board.h>'.
The second category is kept because it was suggested by other maintainers that
simple arch-dependent headers (like for the declaration of a CPU state) should
not imply the inclusion of the whole '<board.h>' and because the architecture
is given also without the former "spec"-mechanism through the type of the build
directory. I think this is questionable but am fine with it.
The third category is kept because the whole way of saying whether
virtualization resp. ARM Trustzone is used is done in an out-dated manner and
changing it now would blow up this commit a lot and exceed the time that I'm
willing to spend. This category should be subject to a future issue.
Ref #4217
The 'src/core/board/<board>/board.h' header is thought as front end of
hardware-specific headers of a given board towards the generic base-hw Core
code. Therefore it leads to problems (circular includes) if the board.h header
is included from within another hardware-specific header.
If hardware-specific headers access declarations from namespace Board in a
definition, the definition should be moved to a compilation unit that may
include board.h. If hardware-specific headers access declarations from board.h
in a declaration, they should either use the primary declaration from the
original header or, if the declaration must be selected according to the board,
another board-specific header should be introduced to reflect this abstraction.
This is applied by this commit for the current state of base-hw.
Ref #4217
It is not necessary to have a class, an object, and a generic header for the
perfomance counter. The kernel merely enables the counter using cpu registers
('msr' instructions, no MMIO) on arm_v6 and arm_v7 only. Therefore this commit
makes the header arm-specific and replaces class and global static object with
a function for enabling the counter.
Fixes#4217
Let the kernel's driver for the global IRQ controller be a member of the one
Kernel::Main object instead of having it as static variables in the drivers for
the local IRQ controllers. Note that this commit spares out renaming 'Pic' to
'Local_interrupt_controller' which would be more sensible now with the new
'Global_interrupt_controller' class. Furthermore, on ARM boards the commit
doesn't move 'Distributer' stuff to the new global IRQ controller class as they
don't have real data members (only MMIO) and can be instanciated for each CPU
anew. However, the right way would be to instanciate them only once in Main as
well.
Ref #4217
The unmanaged-singleton approach was used in this context only because of the
alignment requirement of the Core main-UTCB. This, however can also be achieved
with the new 'Aligned' utility, allowing the UTCB to be a member of the Core
main-thread object.
Ref #4217
It's sufficient to access the boot info only on kernel initialization time.
Therfore, it can remain completely hidden to the rest of the kernel inside
kernel/main.cc in the initialization function.
Ref #4217
This commit introduces the Kernel::Main class that replaces the former way of
initializing the kernel (former 'kernel_init' function) and calling the C++
kernel entry handler (former 'kernel' function). These two are now
'Main::initialize_and_handle_kernel_entry' and 'Main::handle_kernel_entry'.
Also reading the execution time of the idle threads was already moved to
'Main'. The one static Main instance is meant to successivly replace all the
global static objects of the base-hw kernel with data members of the Main
instance making the data model of the kernel much more comprehensible. The
instance and most of its interface are hidden in kernel/main.cc. There are only
rare cases where parts of the Main interface must be accessible from the
outside. This should be done in the most specific way possible (see main.h)
and, if possible, without handing out references to Main data members or the
Main instance itself.
Ref #4217
Normally, the board header can be found for each supported board under
'src/core/board/<BOARD>/board.h'. This was not the case for the board 'pc'
that was located under 'src/core/spec/x86_64/board.h'. The commit fixes this.
Ref #4217
The class name Core_thread in Kernel for the object of the first thread of
core is too generic as there can be an arbitrary number of threads in core
besides this one. Furthermore, creating a core thread has its own syscall
'new_core_thread' that isn't related in any way to Core_thread. Therefore
this commit introduces the more specific name Core_main_thread as replacement
for Core_thread.
Ref #4217
The function was only still used for reading the execution time of idle threads
of CPUs. Certainly, it is technically fine and more performant to read these
values directly from the kernel objects without doing a syscall. However,
calling cpu_pool() for it provides read and write access to a lot more than
only the execution time values. The interface via which Core directly reads
state of the kernel should be as narrow and specific as possible.
Perspectively, we want to get rid of the cpu_pool() accessor anyway. Therefore
this commit introduces Kernel::read_idle_thread_execution_time(cpu_idx) as
replacement. The function is implemented in kernel code and called by Core in
platform.cc.
Ref #4217
Apparently, there is no need for exposing the data members of Trace_source, so,
we sould better make them private before someone gets the impression that they
are meant to be accessed directly.
Ref #4217
Core used to read the kernel-reserved IRQs from the timer objects in the
kernel's CPU objects and the PIC class (inter-processor IRQ). Besides not
being "good style" to access a kernel object in Core, this becomes a problem
when trying to prevent CPU pool from being accessed via global functions.
As a solution, this commit extends the boot info to also carry an array of all
kernel-reserved IRQs.
Ref #4217
For the constructor of Kernel_object<T> there are two variants. One for the
case that it is called from Core where the kernel object (type T) must be
created via a syscall and one when it is called from within the kernel and the
kernel object can be created directly. Selecting one of these variants was done
using a bool argument to the constructor. However, this implies that the
constructor of Kernel_object<T> and that of T have the same signature in the
variadic arguments, even in the syscall case, although technically it would
then not be necessary.
This becomes a problem as soon as kernel objects created by Core shall receive
additional arguments from the kernel, for instance a reference to the global
CPU pool, and therefore stands in the way when wanting to get rid of global
statics in the kernel. Therefore, this commit introduces two constructors that
are selected through enum arguments:
! Kernel_object(Called_from_kernel, ...);
! Kernel_object(Called_from_core, ...);
Ref #4217
The various mapping methods are modelled after the requirements of
the Intel GPUs or rather the Mesa driver back end.
With upcoming support for other driver back ends, we need to
sequeeze their requirements in as well. For now hijack 'map_buffer'
to provide for specifying the kind of attributes the client needs.
For now all buffers mapped in the GGTT for Intel GPUs are treated
as RW.
Issue #4265.
This call allows for checking if the given execution buffer has been
completed and complements the completion signal. Initially the GPU
multiplexer always sent such a signal when the currently scheduled
execution buffer has been completed. During enablement of the 'iris'
driver it became necessary to properly check of sequence number.
In case of the Intel GPU multiplexer the sequence numbers are
continous, which prompted the greater-than-or-equal check in the
DRM back end. By hidding this implementation detail behind the
interface, GPU drivers are free to deal with sequence numbers any
way they like and allows for polling in the client, where the
completion signal is now more of a progress signal.
Issue #4265.
The current info implementation (as RPC) is limited in a few ways:
* The amount of data that may be transferred is constrained by the
underlying base platform
* Most information never changes during run time but is copied
nonetheless
* The information differs depending on the used GPU device and
in its current implementation only contains Intel GPU specific
details
With this commit the 'info' RPC call is replaced with the
'info_dataspace' call that transfers the capability for the dataspace
containing the information only. This is complemented by a client
local 'attached_info' call that allows for getting typed access to
the information. The layout of the information is moved to its own
and GPU-specific header file, e.g., 'gpu/info_intel.h'
Issue #4265.
Rather than using the dataspace capability directly, let the client
choose its own local identifier that is linked to the underlying
capability.
Fixes#4265.
Right now the warning about failure to forward packet from driver to
uplink RX connection reads:
"exception while trying to forward packet from driverto Uplink
connection TX"
Add missing space between "driver" and "to".
Issue #4264
32KB is a rather small value. The driver can cope with it now, but
it does not perform as well as it should. This visible especially
in scenarions like nic_router_flood where we still often hit
synchronous wait path. Bump the size to 256kB.
Issue #4264
The problem can be seen when running nic_router_flood scenarion on arm
qemu_virt boards. With the amount of data this scenario tries to send
the driver quickly complains it has failed to push data into TX VirtIO
queue. After this warning message is printed nothing really happens and
after a while the test scenario fails.
The fact that we can't write all available data to the device is not
unexpected. VirtIO queue size is slected at initialization time and we
don't change it during driver lifetime. It can be tweaked via driver
config, but this does not change the fact that we'll always be able to
produce more data packets than we have free space in the VirtIO queue.
IMO the expected behavior of the driver in such case should be to:
1. Notify the device there is data to process.
2. Wait for the device to process at least part of it.
3. Retry sending queued packets.
One could expect returning Transmit_result::RETRY from _drv_transmit_pkt
would produce such result. Unfortunately it seems that Uplink_client_base
treats RETRY return value as indication of link being down. It'll retry
sending the packet only after the device notifies it the link is once
again up. This is the reason why nothing happens when running
nic_router_flood on top of virtio_nic driver. The link never goes down
in this case so once we fill the TX VirtIO queue and tell the base class
to retry the send, we'll be stuck waiting for link up change event
which will never arrive.
To fix this problem, when sending a packet to the device fails, do a
synchrnonus TX VirtIO queue flush (tell device there is data to process
and wait until its done with it).
With this fix in place nic_router_flood test scenario passes on both arm
qemu_virt boards.
Issue #4264
The contents of those descriptor rings can be modified by the device.
Mark them as volatile so the compiler does not make any assumptions
about them.
Issue #4264
This commit contains features and buf fixes:
* Catch errors during resource allocation
* Because Mesa tries to allocate fence (hardware) registers for each
batch buffer execution, do not allocate new fences for buffer objects
that are already fenced
* Add support for global hardware status page. Each context additionally
has a per-process hardware status page, which we used to set the
global hardware status page during Vgpu switch. This was obviously
wrong. There is only one global hardware status page (set once during
initialization) and a distinct per-process page for contexts.
* Write the sequence number of the currently executing batch buffer to
dword 52 of the per-process hardware status page. We use the pipe line
command with QW_WRITE (quad word write), GLOBAL_GTT_IVB disabled
(address space is per-process address space), and STORE_DATA_INDEX
enabled (write goes to offset of hardware status page). This command
used to write to the scratch page. But Linux now uses the first
reserved word of the per-process hardware status page.
* Add Gen9+ WaEnableGapsTsvCreditFix workaround. This sets the "GAPS TSV
Credit fix Enable" bit of the Arbiter control register (GARBCNTLREG)
as described by the documentation this bit should be set by the BIOS
but is not on most Gen9/9.5 platforms. Not setting this bit leads to
random GPU hangs.
* Increase the context size from 20 to 22 pages for Gen9. On Gen8 the
hardware context is 20 pages (1 hardware status page + 19 ring context
register pages). On Gen9 the size of the ring context registers has
increased by two pages to 21 pages or 81.3125 KBytes as the IGD
documentation states.
* The logical ring size in the ring buffer control of the execlist
context has to be programmed with number of pages - 1. So 0 is 1 page.
We programmed the actual number of pages before, leading to ring
buffer execution of NOOPs if page behind our ring buffer was empty or
GPU hangs if there was data on the page.
issue #4260
* Wait for for completion before return from 'execbuffer2'. This makes
buffer execution synchronous.
* Because the Iris driver manages the virtual address space of the GPU
and creates one GEM context for each batch buffer we have to map/unmap
all buffer objects before and after batch buffer execution.
issue #4260
The lx_emul_virt_to_pages implementation initialized the page ref
counter only for the first page, leaving the remaining elements in
uninitialized state. This, in turn, rendered the Linux page_pool (as
used by the emac network driver) ineffective, ultimately leading the a
memory leak. The fix changes the call of 'init_page_count' to take the
loop variable as argument.
Issue #4225
Increased number of trace subjects since the test sporadically fails on
some platforms.
Also added a sanity check to print an error message in case we run into
the same issue again.
Fixesgenodelabs/genode#4261
This patch adds a switch to internal poll function in libssh that
allows to force this function to immediately return without actually
polling for data and in consequence processing this data. This switch
is used to avoid calling callback functions when flushing output
streams which caused locks due to recursive access to internal
ssh_terminal sessions registry.
Issue #4258
This patch allows to replace sftp packet read and write with
completely asynchronous versions needed to properly hook in existing
ssh_terminal implementation.
Issue #4258
Using 'alignas' in declarations might cause GCC to request for an
implementation of 'operator delete(void*, unsigned long, std::align_val_t)'
although it might actually never be called. This commit adds a dummy
implementation to 'cxx/new_delete.cc' that does nothing more than printing an
error to the log that a proper implementation is missing. This approach is
coherent with our treatment of other global delete operators.
Ref #4217
If one has an object X that has a minimum alignment requirement specified
through 'alignas' this requirement is normally inherited by objects that have
object X as member, and by those that have objects as member that have X as
member, and so on... . However, this chain used to get silently interrupted
(dropping the minimum alignment requirement to 8 again) at objects that are
managed with Genode::Reconstructible or Genode::Constructible. In order to fix
this, the commit ensures that Genode::Reconstructible (and therefore also
Genode::Constructible) has at least the minimum alignment requirement (using
'alignas') as the object it manages.
Ref #4217
The NIC router parses, stores and forwards DNS domain names from DHCP replies.
Yet the routers DHCP client used to not request DNS domain-name information on
DHCP requests. This caused DHCP servers to skip this information on their
replies although it was available. This commit fixes the issue by adding the
DNS domain name code to the request parameter list of requests from the routers
DHCP client.
The 'black_hole' component provides dummy implementations of common
session interfaces.
At this time, only the 'Audio_out' session is provided if enabled
in the configuration of the component:
<config>
<audio_out/>
</config>
Issue #3653
* Switch mesa support from DRI to gallium
Supported drivers are
- softpipe (Sebstian Sumpf)
- iris for Intel GPUs (Alexander Boetcher)
- etnaviv for Vivante GPUs (Josef Söntgen)
* Mesa's generated files are placed into 'contrib/mesa-<hash>/generated'
and are cloned per default from a separate Git repo in order to avoid
hash updates upon package build. In case you need to generate files
yourself use
! prepare_port mesa GENERATE_FILES=1
issue #4254
Support for iris and etnvaviv
* entaviv:
- libdrm on FreeBSD is not prepared for !PCI (and libc our is missing
<sys/pciio.h>
- missing <sys/types.h> include in xf86drmMode.c
- etnaviv relies on linux header files - dummy in $(INC_DIR)
- IOCTL FreeBSD ↔ Linux have swapped IO/OUT bit
- O_CLOEXEC differs between FreeBSD ↔ Linux
issue #4254
According to spec the tail pointer points to the next qword instructions
which will be used by the software.
p 1354, Doc Ref # IHD-OS-BDW-Vol 2c-11.15
issue #4254
Superpages (2M, 1G) are not supported by now, but partially copied over code
from base-hw was around. Remove unused register definitions and remove
non-working super page code do avoid confusion.
issue #4254
Size argument of ggtt free range check is ignored, which leads to
overlapping allocations inside, which leads to unavailable IO-MEM exceptions
thrown by core.
issue #4254
Both, trace_logger and vfs_trace had their own trace_buffer.h. This
commit consolidates the existing implementations and provides the
resulting trace_buffer.h at 'include/trace/'. It thereby becomes part of
the trace api archive.
genodelabs/genode#4244
Driver code such as mfd-core.c may pass 0 as argument n to kcalloc,
which eventually results in an allocation size 0.
res = kcalloc(cell->num_resources, sizeof(*res), GFP_KERNEL);
Since 'res' is checked against NULL for success, kmalloc must not return
a NULL pointer in this case. The patch works around this issue by
forcing an allocation size of 1 byte in this case.
Issue #4253
Clock providers such as drivers/clk/sunxi-ng/ccu-sun8i-r.c don't use
regular init calls but declare their init functions via CLK_OF_DECLARE,
which fill the __clk_of_table. Linux populates the table statically by
using special sections declared in the linker script. In contrast, we
populate the table by expanding the macro to global constructor
functions.
The __clk_of_table is then processed by the call of of_clk_init(NULL).
Issue #4253
* Disable trace source and release ownership on subject destruction.
* Note, since the policy module is also destroyed on descruction of the
session component, the traced component must not access the policy
module when acknowledging the disabled state (else: page fault).
Fixesgenodelabs/genode#4247
If the trace subjects are not properly destructed when the TRACE client
disappears, enabled sources will be owned by a non-existing client.
In other words, when a TRACE client disappears all sources owned by the
client must be disabled.
genodelabs/genode#4247
test-trace always passed, although tracing was never enabled because the
trace subject was not within the first 32 subjects.
* increase number of queried subjects
* output error if trace subject was not found
genodelabs/genode#4247
With this commit, the NIC router DHCP client reads out the first DNS domain
name (DHCP option 15) if any from a DHCP reply that generates an IPv4 config
for a domain and stores the name together with the IPv4 config for that domain.
DNS domain names are reported via the new report tag '<dns-domain>' if the
'config' attribute in the config tag '<report>' is set.
Furthermore, the NIC router DHCP server becomes able to obtain a DNS domain
name from another domain that has a DHCP client dynamically (given the config
attribute 'dns_config_from' is set and no static DNS config is given) or
statically from its configuration (new config tag '<dns-domain>') and propagate
this name with DHCP replies (DHCP option 15).
The 'nic_router_dhcp_*' tests are adapted to test the new feautures.
The commit also gets rid of some mirrored files in
'test/nic_router_dhcp/manager'.
Fixes#4246
WARNING: BREAKS CONFIG COMPATIBILITY!
This commit changes the configuration interface of the NIC router in a way that
may break systems that use the component without proper adjustment!
How to adjust:
At each occurrence of the 'dns_server_from' attribute in a NIC router
configuration replace the attribute name with 'dns_config_from'. The attribute
value remains unaltered.
DETAILED DESCRIPTION
The new attribute name 'dns_config_from' reflects that also other aspects of
the DNS configuration of the denominated domain are used by the DHCP server
that holds the attribute. This commit is a preparation for forwarding also the
domain name (DHCP option 15) with the mechanism behind the attribute.
Ref #4246
The fact that the IPv4 config was a struct with all data members public was a
mere leftover of an early state of the NIC router. Today, the router
implementation style is to avoid structs and public data members wherever
possible.
This commit slightly changes the behavior of the router regarding log output.
The router used to print malformed IPv4 configurations to the log only if
the 'verbose' config flag was set using this style:
! [my_domain] malformed dynamic IP config: interface 10.0.2.1/24 ...
Now, malformed IPv4 configurations are only printed if the
'verbose_domain_state' config flag is set (like with any IP4v configuration
states) using this style:
! [my_domain] dynamic IP config: malformed (interface 10.0.2.1/24 ...)
Fixes#4242
The NIC router DHCP server used to add an extra option 6 field to DHCP replies
for each DNS server address. This conflicts with RFC #2132 section 3.8 which
states that the addresses should be listed within one option 6 field without
delimiter. The discrepancy is fixed by this commit.
Ref #4242
File size must be the same as the number of bytes that can be read from
the file. Otherwise, this will trigger a `Truncated_during_read`
exception.
Fixesgenodelabs/genode#4240
Via a new configuration attribute, the user can decide whether the router
should answer dropped fragmented IPv4 with an ICMP "destination unreachable"
packet and, if so, which value the ICMP code field of this packet should have.
The default is that the router doesn't send such responses (silently dropping
fragmented IPv4). The behavior is tested by the 'nic_router_ipv4_fragm' test.
Fixes#4236
If the new attribute 'dropped_fragm_ipv4' of the <report> tag in the NIC router
config is set "yes", the router will report the number of packets that were
dropped per interface respectively domain because fragmented IPv4 is not
supported. The default is not to report the counter. The behavior is tested by
the 'nic_router_ipv4_fragm' test.
Ref #4236
The NIC router used to ignore the IPv4 header fields "More fragments" and
"Fragment offset" completely. Therefore higher-level protocols of fragmented
IPv4 were interpreted wrong because each fragment was considered a self-
standing packet, expecting, for instance UDP/TCP headers somewhere inside of
the UDP/TCP data field. Normally, such packets were dropped as soon as the
UDP/TCP checksum check failed because of the misinterpretation. However,
it was also possible for fragmented IPv4 to pass the router although normally
only partially.
IPv4 fragmentation support in the router would introduce some potential
security risks and is presumably not an easy endeavor. So, for now, we settled
on not supporting IPv4 fragmentation. With this commit, the router simply drops
all fragmented IPv4. This is reflected to the log for each fragment as "drop
packet (fragmented IPv4 not supported)" when 'verbose_packet_drop="yes"' is
configured.
The new test 'run/nic_router_ipv4_fragm' is an automated test for this
behavior. The test is added to the autopilot list.
Ref #4236
