mirrors/serenity
1
0
Fork 0
mirror of https://github.com/SerenityOS/serenity.git synced 2025-01-22 17:31:58 -05:00
Code Issues Projects Releases Packages Wiki Activity
serenity/Kernel/Lock.cpp

344 lines
13 KiB
C++
Raw Normal View History

Meta: Add license header to source files As suggested by Joshua, this commit adds the 2-clause BSD license as a comment block to the top of every source file. For the first pass, I've just added myself for simplicity. I encourage everyone to add themselves as copyright holders of any file they've added or modified in some significant way. If I've added myself in error somewhere, feel free to replace it with the appropriate copyright holder instead. Going forward, all new source files should include a license header.
2020-01-18 03:38:21 -05:00
/*
* Copyright (c) 2018-2020, Andreas Kling <kling@serenityos.org>
*
Everything: Move to SPDX license identifiers in all files. SPDX License Identifiers are a more compact / standardized way of representing file license information. See: https://spdx.dev/resources/use/#identifiers This was done with the `ambr` search and replace tool. ambr --no-parent-ignore --key-from-file --rep-from-file key.txt rep.txt *
2021-04-22 04:24:48 -04:00
* SPDX-License-Identifier: BSD-2-Clause
Meta: Add license header to source files As suggested by Joshua, this commit adds the 2-clause BSD license as a comment block to the top of every source file. For the first pass, I've just added myself for simplicity. I encourage everyone to add themselves as copyright holders of any file they've added or modified in some significant way. If I've added myself in error somewhere, feel free to replace it with the appropriate copyright holder instead. Going forward, all new source files should include a license header.
2020-01-18 03:38:21 -05:00
*/
Kernel: Fix signal delivery When delivering urgent signals to the current thread we need to check if we should be unblocked, and if not we need to yield to another process. We also need to make sure that we suppress context switches during Process::exec() so that we don't clobber the registers that it sets up (eip mainly) by a context switch. To be able to do that we add the concept of a critical section, which are similar to Process::m_in_irq but different in that they can be requested at any time. Calls to Scheduler::yield and Scheduler::donate_to will return instantly without triggering a context switch, but the processor will then asynchronously trigger a context switch once the critical section is left.
2020-07-03 07:19:50 -04:00
#include <AK/TemporaryChange.h>
Meta: Split debug defines into multiple headers. The following script was used to make these changes: #!/bin/bash set -e tmp=$(mktemp -d) echo "tmp=$tmp" find Kernel \( -name '*.cpp' -o -name '*.h' \) | sort > $tmp/Kernel.files find . \( -path ./Toolchain -prune -o -path ./Build -prune -o -path ./Kernel -prune \) -o \( -name '*.cpp' -o -name '*.h' \) -print | sort > $tmp/EverythingExceptKernel.files cat $tmp/Kernel.files | xargs grep -Eho '[A-Z0-9_]+_DEBUG' | sort | uniq > $tmp/Kernel.macros cat $tmp/EverythingExceptKernel.files | xargs grep -Eho '[A-Z0-9_]+_DEBUG' | sort | uniq > $tmp/EverythingExceptKernel.macros comm -23 $tmp/Kernel.macros $tmp/EverythingExceptKernel.macros > $tmp/Kernel.unique comm -1 $tmp/Kernel.macros $tmp/EverythingExceptKernel.macros > $tmp/EverythingExceptKernel.unique cat $tmp/Kernel.unique | awk '{ print "#cmakedefine01 "$1 }' > $tmp/Kernel.header cat $tmp/EverythingExceptKernel.unique | awk '{ print "#cmakedefine01 "$1 }' > $tmp/EverythingExceptKernel.header for macro in $(cat $tmp/Kernel.unique) do cat $tmp/Kernel.files | xargs grep -l $macro >> $tmp/Kernel.new-includes ||: done cat $tmp/Kernel.new-includes | sort > $tmp/Kernel.new-includes.sorted for macro in $(cat $tmp/EverythingExceptKernel.unique) do cat $tmp/Kernel.files | xargs grep -l $macro >> $tmp/Kernel.old-includes ||: done cat $tmp/Kernel.old-includes | sort > $tmp/Kernel.old-includes.sorted comm -23 $tmp/Kernel.new-includes.sorted $tmp/Kernel.old-includes.sorted > $tmp/Kernel.includes.new comm -13 $tmp/Kernel.new-includes.sorted $tmp/Kernel.old-includes.sorted > $tmp/Kernel.includes.old comm -12 $tmp/Kernel.new-includes.sorted $tmp/Kernel.old-includes.sorted > $tmp/Kernel.includes.mixed for file in $(cat $tmp/Kernel.includes.new) do sed -i -E 's/#include <AK\/Debug\.h>/#include <Kernel\/Debug\.h>/' $file done for file in $(cat $tmp/Kernel.includes.mixed) do echo "mixed include in $file, requires manual editing." done
2021-01-25 10:07:10 -05:00
#include <Kernel/Debug.h>
Kernel: Move all code into the Kernel namespace
2020-02-15 19:27:42 -05:00
#include <Kernel/KSyms.h>
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
#include <Kernel/Lock.h>
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
#include <Kernel/Thread.h>
Kernel: Move all code into the Kernel namespace
2020-02-15 19:27:42 -05:00
namespace Kernel {
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
void Lock::lock(Mode mode)
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
{
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
lock("unknown", 0, mode);
Kernel: Introduce shared locking mode A Lock can now be held either in shared or exclusive mode. Multiple threads can hold the same lock in shared mode at one time, but if any thread holds the lock in exclusive mode, no other thread can hold it at the same time in either mode.
2020-04-18 05:21:00 -04:00
}
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
void Lock::lock(const char* file, int line, Mode mode)
#else
Kernel: Introduce shared locking mode A Lock can now be held either in shared or exclusive mode. Multiple threads can hold the same lock in shared mode at one time, but if any thread holds the lock in exclusive mode, no other thread can hold it at the same time in either mode.
2020-04-18 05:21:00 -04:00
void Lock::lock(Mode mode)
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
#endif
Kernel: Introduce shared locking mode A Lock can now be held either in shared or exclusive mode. Multiple threads can hold the same lock in shared mode at one time, but if any thread holds the lock in exclusive mode, no other thread can hold it at the same time in either mode.
2020-04-18 05:21:00 -04:00
{
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
// NOTE: This may be called from an interrupt handler (not an IRQ handler)
// and also from within critical sections!
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!Processor::current().in_irq());
VERIFY(mode != Mode::Unlocked);
Kernel: Turn Thread::current and Process::current into functions This allows us to query the current thread and process on a per processor basis
2020-06-28 17:34:31 -04:00
auto current_thread = Thread::current();
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
ScopedCritical critical; // in case we're not in a critical section already
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
for (;;) {
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
if (m_lock.exchange(true, AK::memory_order_acq_rel) != false) {
// I don't know *who* is using "m_lock", so just yield.
Scheduler::yield_from_critical();
continue;
}
// FIXME: Do not add new readers if writers are queued.
Mode current_mode = m_mode;
switch (current_mode) {
case Mode::Unlocked: {
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_TRACE_DEBUG, "Lock::lock @ ({}) {}: acquire {}, currently unlocked", this, m_name, mode_to_string(mode));
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_mode = mode;
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!m_holder);
VERIFY(m_shared_holders.is_empty());
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
if (mode == Mode::Exclusive) {
m_holder = current_thread;
} else {
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(mode == Mode::Shared);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_shared_holders.set(current_thread, 1);
}
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_times_locked == 0);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_times_locked++;
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix LOCK_DEBUG feature to work again - UBSAN detected cases where we were calling thread->holding_lock(..) but current_thread was nullptr. - Fix Lock::force_unlock_if_locked to not pass the correct ref delta to holding_lock(..).
2021-04-24 17:16:31 -04:00
if (current_thread) {
current_thread->holding_lock(*this, 1, file, line);
}
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
#endif
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
m_queue.should_block(true);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_lock.store(false, AK::memory_order_release);
return;
}
case Mode::Exclusive: {
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_holder);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
if (m_holder != current_thread)
break;
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_shared_holders.is_empty());
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
if constexpr (LOCK_TRACE_DEBUG) {
if (mode == Mode::Exclusive)
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
dbgln("Lock::lock @ {} ({}): acquire {}, currently exclusive, holding: {}", this, m_name, mode_to_string(mode), m_times_locked);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
else
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
dbgln("Lock::lock @ {} ({}): acquire exclusive (requested {}), currently exclusive, holding: {}", this, m_name, mode_to_string(mode), m_times_locked);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
}
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(mode == Mode::Exclusive || mode == Mode::Shared);
VERIFY(m_times_locked > 0);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_times_locked++;
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
current_thread->holding_lock(*this, 1, file, line);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
#endif
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_lock.store(false, AK::memory_order_release);
return;
}
case Mode::Shared: {
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!m_holder);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
if (mode != Mode::Shared)
break;
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_TRACE_DEBUG, "Lock::lock @ {} ({}): acquire {}, currently shared, locks held {}", this, m_name, mode_to_string(mode), m_times_locked);
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_times_locked > 0);
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_times_locked++;
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!m_shared_holders.is_empty());
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
auto it = m_shared_holders.find(current_thread);
if (it != m_shared_holders.end())
it->value++;
else
m_shared_holders.set(current_thread, 1);
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
current_thread->holding_lock(*this, 1, file, line);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
#endif
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_lock.store(false, AK::memory_order_release);
return;
}
default:
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY_NOT_REACHED();
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
}
Kernel: Fix race condition in Lock::lock that may leave corrupted state
2021-01-23 12:43:52 -05:00
m_lock.store(false, AK::memory_order_release);
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_TRACE_DEBUG, "Lock::lock @ {} ({}) waiting...", this, m_name);
Kernel: Add WaitQueue::wait_forever and it use it for all infinite waits. In preparation for marking BlockingResult [[nodiscard]], there are a few places that perform infinite waits, which we never observe the result of the wait. Instead of suppressing them, add an alternate function which returns void when performing and infinite wait.
2021-02-14 18:02:14 -05:00
m_queue.wait_forever(m_name);
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_TRACE_DEBUG, "Lock::lock @ {} ({}) waited", this, m_name);
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
}
}
void Lock::unlock()
{
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
// NOTE: This may be called from an interrupt handler (not an IRQ handler)
// and also from within critical sections!
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!Processor::current().in_irq());
Kernel: Turn Thread::current and Process::current into functions This allows us to query the current thread and process on a per processor basis
2020-06-28 17:34:31 -04:00
auto current_thread = Thread::current();
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
ScopedCritical critical; // in case we're not in a critical section already
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
for (;;) {
Kernel: Minor Lock optimization
2020-10-28 22:03:15 -04:00
if (m_lock.exchange(true, AK::memory_order_acq_rel) == false) {
Kernel: Specify default memory order for some non-synchronizing Atomics
2021-01-03 18:58:50 -05:00
Mode current_mode = m_mode;
Everywhere: Debug macros instead of constexpr. This was done with the following script: find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/dbgln<debug_([a-z_]+)>/dbgln<\U\1_DEBUG>/' {} \; find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/if constexpr \(debug_([a-z0-9_]+)/if constexpr \(\U\1_DEBUG/' {} \;
2021-01-23 17:59:27 -05:00
if constexpr (LOCK_TRACE_DEBUG) {
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
if (current_mode == Mode::Shared)
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
dbgln("Lock::unlock @ {} ({}): release {}, locks held: {}", this, m_name, mode_to_string(current_mode), m_times_locked);
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
else
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
dbgln("Lock::unlock @ {} ({}): release {}, holding: {}", this, m_name, mode_to_string(current_mode), m_times_locked);
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
}
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(current_mode != Mode::Unlocked);
Kernel: Introduce shared locking mode A Lock can now be held either in shared or exclusive mode. Multiple threads can hold the same lock in shared mode at one time, but if any thread holds the lock in exclusive mode, no other thread can hold it at the same time in either mode.
2020-04-18 05:21:00 -04:00
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_times_locked > 0);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_times_locked--;
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
switch (current_mode) {
case Mode::Exclusive:
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_holder == current_thread);
VERIFY(m_shared_holders.is_empty());
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
if (m_times_locked == 0)
m_holder = nullptr;
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
break;
case Mode::Shared: {
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!m_holder);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
auto it = m_shared_holders.find(current_thread);
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(it != m_shared_holders.end());
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
if (it->value > 1) {
it->value--;
} else {
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(it->value > 0);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_shared_holders.remove(it);
}
break;
}
default:
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY_NOT_REACHED();
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
}
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
bool unlocked_last = (m_times_locked == 0);
if (unlocked_last) {
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(current_mode == Mode::Exclusive ? !m_holder : m_shared_holders.is_empty());
Kernel: Specify default memory order for some non-synchronizing Atomics
2021-01-03 18:58:50 -05:00
m_mode = Mode::Unlocked;
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
m_queue.should_block(false);
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
}
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix LOCK_DEBUG feature to work again - UBSAN detected cases where we were calling thread->holding_lock(..) but current_thread was nullptr. - Fix Lock::force_unlock_if_locked to not pass the correct ref delta to holding_lock(..).
2021-04-24 17:16:31 -04:00
if (current_thread) {
current_thread->holding_lock(*this, -1);
}
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
#endif
Kernel: Introduce shared locking mode A Lock can now be held either in shared or exclusive mode. Multiple threads can hold the same lock in shared mode at one time, but if any thread holds the lock in exclusive mode, no other thread can hold it at the same time in either mode.
2020-04-18 05:21:00 -04:00
Kernel: Fix some issues related to fixes and block conditions Fix some problems with join blocks where the joining thread block condition was added twice, which lead to a crash when trying to unblock that condition a second time. Deferred block condition evaluation by File objects were also not properly keeping the File object alive, which lead to some random crashes and corruption problems. Other problems were caused by the fact that the Queued state didn't handle signals/interruptions consistently. To solve these issues we remove this state entirely, along with Thread::wait_on and change the WaitQueue into a BlockCondition instead. Also, deliver signals even if there isn't going to be a context switch to another thread. Fixes #4336 and #4330
2020-12-07 23:29:41 -05:00
m_lock.store(false, AK::memory_order_release);
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
if (unlocked_last) {
u32 did_wake = m_queue.wake_one();
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_TRACE_DEBUG, "Lock::unlock @ {} ({}) wake one ({})", this, m_name, did_wake);
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
}
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
return;
}
Kernel: Add a WaitQueue for Thread queueing/waking and use it for Lock The kernel's Lock class now uses a proper wait queue internally instead of just having everyone wake up regularly to try to acquire the lock. We also keep the donation mechanism, so that whenever someone tries to take the lock and fails, that thread donates the remainder of its timeslice to the current lock holder. After unlocking a Lock, the unlocking thread calls WaitQueue::wake_one, which unblocks the next thread in queue.
2019-12-01 05:57:20 -05:00
// I don't know *who* is using "m_lock", so just yield.
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
Scheduler::yield_from_critical();
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
}
}
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
auto Lock::force_unlock_if_locked(u32& lock_count_to_restore) -> Mode
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
{
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
// NOTE: This may be called from an interrupt handler (not an IRQ handler)
// and also from within critical sections!
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!Processor::current().in_irq());
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
auto current_thread = Thread::current();
ScopedCritical critical; // in case we're not in a critical section already
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
for (;;) {
if (m_lock.exchange(true, AK::memory_order_acq_rel) == false) {
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
Mode previous_mode;
auto current_mode = m_mode.load(AK::MemoryOrder::memory_order_relaxed);
switch (current_mode) {
case Mode::Exclusive: {
if (m_holder != current_thread) {
m_lock.store(false, AK::MemoryOrder::memory_order_release);
lock_count_to_restore = 0;
return Mode::Unlocked;
}
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_RESTORE_DEBUG, "Lock::force_unlock_if_locked @ {}: unlocking exclusive with lock count: {}", this, m_times_locked);
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
#if LOCK_DEBUG
Kernel: Fix LOCK_DEBUG feature to work again - UBSAN detected cases where we were calling thread->holding_lock(..) but current_thread was nullptr. - Fix Lock::force_unlock_if_locked to not pass the correct ref delta to holding_lock(..).
2021-04-24 17:16:31 -04:00
m_holder->holding_lock(*this, -(int)m_times_locked);
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
#endif
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_holder = nullptr;
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_times_locked > 0);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
lock_count_to_restore = m_times_locked;
m_times_locked = 0;
Kernel: Specify default memory order for some non-synchronizing Atomics
2021-01-03 18:58:50 -05:00
m_mode = Mode::Unlocked;
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
m_queue.should_block(false);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_lock.store(false, AK::memory_order_release);
previous_mode = Mode::Exclusive;
break;
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
}
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
case Mode::Shared: {
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!m_holder);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
auto it = m_shared_holders.find(current_thread);
if (it == m_shared_holders.end()) {
m_lock.store(false, AK::MemoryOrder::memory_order_release);
lock_count_to_restore = 0;
return Mode::Unlocked;
}
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_RESTORE_DEBUG, "Lock::force_unlock_if_locked @ {}: unlocking exclusive with lock count: {}, total locks: {}",
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
this, it->value, m_times_locked);
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(it->value > 0);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
lock_count_to_restore = it->value;
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(lock_count_to_restore > 0);
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_holder->holding_lock(*this, -(int)lock_count_to_restore);
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
#endif
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_shared_holders.remove(it);
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_times_locked >= lock_count_to_restore);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_times_locked -= lock_count_to_restore;
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
if (m_times_locked == 0) {
Kernel: Specify default memory order for some non-synchronizing Atomics
2021-01-03 18:58:50 -05:00
m_mode = Mode::Unlocked;
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
m_queue.should_block(false);
}
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_lock.store(false, AK::memory_order_release);
previous_mode = Mode::Shared;
break;
}
case Mode::Unlocked: {
m_lock.store(false, AK::memory_order_relaxed);
lock_count_to_restore = 0;
previous_mode = Mode::Unlocked;
break;
}
default:
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY_NOT_REACHED();
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
}
Kernel: Fix some issues related to fixes and block conditions Fix some problems with join blocks where the joining thread block condition was added twice, which lead to a crash when trying to unblock that condition a second time. Deferred block condition evaluation by File objects were also not properly keeping the File object alive, which lead to some random crashes and corruption problems. Other problems were caused by the fact that the Queued state didn't handle signals/interruptions consistently. To solve these issues we remove this state entirely, along with Thread::wait_on and change the WaitQueue into a BlockCondition instead. Also, deliver signals even if there isn't going to be a context switch to another thread. Fixes #4336 and #4330
2020-12-07 23:29:41 -05:00
m_queue.wake_one();
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
return previous_mode;
}
// I don't know *who* is using "m_lock", so just yield.
Scheduler::yield_from_critical();
}
}
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
void Lock::restore_lock(Mode mode, u32 lock_count)
{
return restore_lock("unknown", 0, mode, lock_count);
}
void Lock::restore_lock(const char* file, int line, Mode mode, u32 lock_count)
#else
void Lock::restore_lock(Mode mode, u32 lock_count)
#endif
{
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(mode != Mode::Unlocked);
VERIFY(lock_count > 0);
VERIFY(!Processor::current().in_irq());
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
auto current_thread = Thread::current();
ScopedCritical critical; // in case we're not in a critical section already
for (;;) {
if (m_lock.exchange(true, AK::memory_order_acq_rel) == false) {
switch (mode) {
case Mode::Exclusive: {
auto expected_mode = Mode::Unlocked;
Kernel: Specify default memory order for some non-synchronizing Atomics
2021-01-03 18:58:50 -05:00
if (!m_mode.compare_exchange_strong(expected_mode, Mode::Exclusive))
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
break;
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_RESTORE_DEBUG, "Lock::restore_lock @ {}: restoring {} with lock count {}, was unlocked", this, mode_to_string(mode), lock_count);
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_times_locked == 0);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_times_locked = lock_count;
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!m_holder);
VERIFY(m_shared_holders.is_empty());
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_holder = current_thread;
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
m_queue.should_block(true);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_lock.store(false, AK::memory_order_release);
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_holder->holding_lock(*this, (int)lock_count, file, line);
#endif
return;
}
case Mode::Shared: {
auto expected_mode = Mode::Unlocked;
Kernel: Specify default memory order for some non-synchronizing Atomics
2021-01-03 18:58:50 -05:00
if (!m_mode.compare_exchange_strong(expected_mode, Mode::Shared) && expected_mode != Mode::Shared)
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
break;
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
Everywhere: Replace dbgln<flag>(...) with dbgln_if(flag, ...) Replacement made by `find Kernel Userland -name '*.h' -o -name '*.cpp' | sed -i -Ee 's/dbgln\b<(\w+)>\(/dbgln_if(\1, /g'`
2021-02-07 07:03:24 -05:00
dbgln_if(LOCK_RESTORE_DEBUG, "Lock::restore_lock @ {}: restoring {} with lock count {}, was {}",
Everywhere: Replace a bundle of dbg with dbgln. These changes are arbitrarily divided into multiple commits to make it easier to find potentially introduced bugs with git bisect.
2021-01-18 11:25:44 -05:00
this, mode_to_string(mode), lock_count, mode_to_string(expected_mode));
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(expected_mode == Mode::Shared || m_times_locked == 0);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_times_locked += lock_count;
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(!m_holder);
VERIFY((expected_mode == Mode::Unlocked) == m_shared_holders.is_empty());
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
auto set_result = m_shared_holders.set(current_thread, lock_count);
// There may be other shared lock holders already, but we should not have an entry yet
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(set_result == AK::HashSetResult::InsertedNewEntry);
Kernel: Fix some race conditions with Lock and waiting/waking threads There is a window between acquiring/releasing the lock with the atomic variables and subsequently waiting or waking threads. With a single processor this window was closed by using a critical section, but this doesn't prevent other processors from running these code paths. To solve this, set a flag in the WaitQueue while holding m_lock which determines if threads should be blocked at all.
2021-01-24 00:30:10 -05:00
m_queue.should_block(true);
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_lock.store(false, AK::memory_order_release);
Everywhere: Use CMake to generate AK/Debug.h. This was done with the help of several scripts, I dump them here to easily find them later: awk '/#ifdef/ { print "#cmakedefine01 "$2 }' AK/Debug.h.in for debug_macro in $(awk '/#ifdef/ { print $2 }' AK/Debug.h.in) do find . \( -name '*.cpp' -o -name '*.h' -o -name '*.in' \) -not -path './Toolchain/*' -not -path './Build/*' -exec sed -i -E 's/#ifdef '$debug_macro'/#if '$debug_macro'/' {} \; done # Remember to remove WRAPPER_GERNERATOR_DEBUG from the list. awk '/#cmake/ { print "set("$2" ON)" }' AK/Debug.h.in
2021-01-23 17:29:11 -05:00
#if LOCK_DEBUG
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
m_holder->holding_lock(*this, (int)lock_count, file, line);
#endif
return;
}
default:
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY_NOT_REACHED();
Kernel: Fix Lock race causing infinite spinning between two threads We need to account for how many shared lock instances the current thread owns, so that we can properly release such references when yielding execution. We also need to release the process lock when donating.
2020-12-14 18:36:22 -05:00
}
m_lock.store(false, AK::memory_order_relaxed);
Kernel: Fix some problems with Thread::wait_on and Lock This changes the Thread::wait_on function to not enable interrupts upon leaving, which caused some problems with page fault handlers and in other situations. It may now be called from critical sections, with interrupts enabled or disabled, and returns to the same state. This also requires some fixes to Lock. To aid debugging, a new define LOCK_DEBUG is added that enables checking for Lock leaks upon finalization of a Thread.
2020-11-30 21:04:36 -05:00
}
// I don't know *who* is using "m_lock", so just yield.
Scheduler::yield_from_critical();
}
Kernel: Oops, forgot to add Lock.cpp.
2019-07-29 06:00:14 -04:00
}
Kernel: Move all code into the Kernel namespace
2020-02-15 19:27:42 -05:00
Kernel: Allow process with multiple threads to call exec and exit This allows a process wich has more than 1 thread to call exec, even from a thread. This kills all the other threads, but it won't wait for them to finish, just makes sure that they are not in a running/runable state. In the case where a thread does exec, the new program PID will be the thread TID, to keep the PID == TID in the new process. This introduces a new function inside the Process class, kill_threads_except_self which is called on exit() too (exit with multiple threads wasn't properly working either). Inside the Lock class, there is the need for a new function, clear_waiters, which removes all the waiters from the Process::big_lock. This is needed since after a exit/exec, there should be no other threads waiting for this lock, the threads should be simply killed. Only queued threads should wait for this lock at this point, since blocked threads are handled in set_should_die.
2020-02-18 07:28:28 -05:00
void Lock::clear_waiters()
{
Everywhere: Rename ASSERT => VERIFY (...and ASSERT_NOT_REACHED => VERIFY_NOT_REACHED) Since all of these checks are done in release builds as well, let's rename them to VERIFY to prevent confusion, as everyone is used to assertions being compiled out in release. We can introduce a new ASSERT macro that is specifically for debug checks, but I'm doing this wholesale conversion first since we've accumulated thousands of these already, and it's not immediately obvious which ones are suitable for ASSERT.
2021-02-23 14:42:32 -05:00
VERIFY(m_mode != Mode::Shared);
Kernel: Fix some issues related to fixes and block conditions Fix some problems with join blocks where the joining thread block condition was added twice, which lead to a crash when trying to unblock that condition a second time. Deferred block condition evaluation by File objects were also not properly keeping the File object alive, which lead to some random crashes and corruption problems. Other problems were caused by the fact that the Queued state didn't handle signals/interruptions consistently. To solve these issues we remove this state entirely, along with Thread::wait_on and change the WaitQueue into a BlockCondition instead. Also, deliver signals even if there isn't going to be a context switch to another thread. Fixes #4336 and #4330
2020-12-07 23:29:41 -05:00
m_queue.wake_all();
Kernel: Allow process with multiple threads to call exec and exit This allows a process wich has more than 1 thread to call exec, even from a thread. This kills all the other threads, but it won't wait for them to finish, just makes sure that they are not in a running/runable state. In the case where a thread does exec, the new program PID will be the thread TID, to keep the PID == TID in the new process. This introduces a new function inside the Process class, kill_threads_except_self which is called on exit() too (exit with multiple threads wasn't properly working either). Inside the Lock class, there is the need for a new function, clear_waiters, which removes all the waiters from the Process::big_lock. This is needed since after a exit/exec, there should be no other threads waiting for this lock, the threads should be simply killed. Only queued threads should wait for this lock at this point, since blocked threads are handled in set_should_die.
2020-02-18 07:28:28 -05:00
}
Kernel: Move all code into the Kernel namespace
2020-02-15 19:27:42 -05:00
}
Reference in a new issue Copy permalink