Re: [PATCH] kern: simple futex for gnumach (version 3)
On 22.12.2013 01:28:37, Richard Braun wrote: > On Sat, Dec 21, 2013 at 11:29:34PM +0100, Marin Ramesa wrote: > > On 21.12.2013 23:20:43, Richard Braun wrote: > > > How about adding everything necessary to actually test it from > > > userspace ? > > > > Sure, what needs to be added? I don't have the conditions to test > > this. I earlier wrote some proof-of-concept stuff that might be > > used as a test, but I don't know how. > > Learn about MiG and how to write RPCs. There is a lot you can find on > the wiki and through search engines. If possible, don't create an > ioctl-style of call, with many possible behaviours through the same > entry point, but make several distinct RPCs instead, then add them to > include/mach/gnumach.defs. Look at how it's done for existing RPCs, > but make sure you read and understand the documentation about MiG and > Mach IPC if you want to work on such low-level stuff. OK, I'm reading the documentation. In the meantime I have defined several simple RPCs for testing purposes. I will send the updated patch shortly.
[PATCH] kern: simple futex for gnumach (version 4)
GPL license included, more bugs fixed and RPCs defined.
---
Makefrag.am | 2 +
include/mach/gnumach.defs | 26 +++
kern/futex.c | 405 ++
kern/futex.h | 108 +
4 files changed, 541 insertions(+)
create mode 100644 kern/futex.c
create mode 100644 kern/futex.h
diff --git a/Makefrag.am b/Makefrag.am
index c1387bd..e43f882 100644
--- a/Makefrag.am
+++ b/Makefrag.am
@@ -145,6 +145,8 @@ libkernel_a_SOURCES += \
kern/eventcount.h \
kern/exception.c \
kern/exception.h \
+ kern/futex.c \
+ kern/futex.h \
kern/host.c \
kern/host.h \
kern/ipc_host.c \
diff --git a/include/mach/gnumach.defs b/include/mach/gnumach.defs
index 12c4e99..8c5ea4b 100644
--- a/include/mach/gnumach.defs
+++ b/include/mach/gnumach.defs
@@ -63,3 +63,29 @@ simpleroutine thread_terminate_release(
reply_port : mach_port_name_t;
address : vm_address_t;
size: vm_size_t);
+
+/*
+ * When operation equals FUTEX_WAIT, this is a method for a thread to wait for
a
+ * change of value at a given address.
+ *
+ * When operation equals FUTEX_WAKE, this is a method for waking up a number
+ * (specified by the argument value) of threads waiting for a change of value
+ * at a given address (threads that are in FUTEX_WAIT).
+ */
+
+simpleroutine futex_rpc(
+ thread : thread_t;
+ address : pointer_t;
+ value : int;
+ operation : int);
+
+simpleroutine futex_wait_rpc(
+ thread : thread_t;
+ address : pointer_t;
+ value : int);
+
+simpleroutine futex_wake_rpc(
+ thread : thread_t;
+ address : pointer_t;
+ thread_num : int);
+
diff --git a/kern/futex.c b/kern/futex.c
new file mode 100644
index 000..7860265
--- /dev/null
+++ b/kern/futex.c
@@ -0,0 +1,405 @@
+/*
+ * Copyright (C) 2013 Free Software Foundation, Inc.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ * Author: Marin Ramesa
+ */
+
+/*
+ * Fast userspace locking
+ *
+ */
+
+#include
+#include
+
+static futex_hash_table_t table = NULL;
+static unsigned int prev_hash_val = 0;
+
+/*
+ * When operation equals FUTEX_WAIT, this is a method for a thread to wait for
a
+ * change of value at a given address.
+ *
+ * When operation equals FUTEX_WAKE, this is a method for waking up a number
+ * (specified by the argument value) of threads waiting for a change of value
+ * at a given address (threads that are in FUTEX_WAIT).
+ *
+ * Return values:
+ * FUTEX_RESUMED_BY_WAKE - thread has been resumed by FUTEX_WAKE
+ * FUTEX_NO_THREAD_SUSPENDED - value at the address has been changed while
inside
+ * FUTEX_WAIT and no thread has been suspended
+ * FUTEX_SOME_NUMBER_RESUMED - value threads have been resumed by
FUTEX_WAKE
+ * FUTEX_NO_THREAD - FUTEX_WAKE did not found a suspended thread at the
passed address
+ * FUTEX_EXISTS - futex already exists at the new address
+ * FUTEX_MEMORY_ERROR - memory error
+ * FUTEX_UNKNOWN_OPERATION - unknown operation
+ * FUTEX_RESOURCE_SHORTAGE - no virtual memory or the mutex is not
per-process
+ *
+ */
+int futex(int *address, int value, int operation)
+{
+ switch (operation) {
+ case FUTEX_WAIT:
+ return futex_wait(address, value);
+ case FUTEX_WAKE:
+ return futex_wake(address, value);
+ default:
+ return FUTEX_UNKNOWN_OPERATION;
+ }
+}
+
+int futex_init(int *address, futex_t futex, futex_hash_table_t hash_table,
unsigned int hash_val)
+{
+ simple_lock_init(&futex->futex_wait_lock);
+
+ if (hash_table == NULL) {
+ hash_table = futex_hash_table_init(INITIAL_HASH_TABLE_SIZE,
current_thread()->task);
+ if (hash_table == NULL)
+ return FUTEX_RESOURCE_SHORTAGE;
+ }
+
+ if ((hash_table->futex_elements =
+ (futex_t)kalloc((hash_table->size + 1) * sizeof(struct futex)))
== NULL) {
+
+ return FUTEX_RESOURCE_SHORTAGE
Re: [PATCH] kern: simple futex for gnumach (version 3)
On Sun, Dec 22, 2013 at 03:32:42PM +0100, Marin Ramesa wrote: > On 22.12.2013 01:28:37, Richard Braun wrote: > > On Sat, Dec 21, 2013 at 11:29:34PM +0100, Marin Ramesa wrote: > > > On 21.12.2013 23:20:43, Richard Braun wrote: > > > > How about adding everything necessary to actually test it from > > > > userspace ? > > > > > > Sure, what needs to be added? I don't have the conditions to test > > > this. I earlier wrote some proof-of-concept stuff that might be > > > used as a test, but I don't know how. > > > > Learn about MiG and how to write RPCs. There is a lot you can find on > > the wiki and through search engines. If possible, don't create an > > ioctl-style of call, with many possible behaviours through the same > > entry point, but make several distinct RPCs instead, then add them to > > include/mach/gnumach.defs. Look at how it's done for existing RPCs, > > but make sure you read and understand the documentation about MiG and > > Mach IPC if you want to work on such low-level stuff. > > OK, I'm reading the documentation. In the meantime I have defined > several simple RPCs for testing purposes. I will send the updated patch > shortly. Test it yourself from userspace before you resubmit. -- Richard Braun
Re: [PATCH] kern: simple futex for gnumach (version 3)
On 22.12.2013 17:56:32, Richard Braun wrote: > > OK, I'm reading the documentation. In the meantime I have defined > > several simple RPCs for testing purposes. I will send the updated > > patch shortly. > > Test it yourself from userspace before you resubmit. That's the problem. I don't know how to test this. How to boot up a modified gnumach and then execute a program over it? I tried with QEMU, but I can't get it to work.
Re: [PATCH] kern: simple futex for gnumach (version 3)
On Sun, Dec 22, 2013 at 06:22:03PM +0100, Marin Ramesa wrote: > On 22.12.2013 17:56:32, Richard Braun wrote: > > Test it yourself from userspace before you resubmit. > > That's the problem. I don't know how to test this. How to boot up a > modified gnumach and then execute a program over it? I tried with QEMU, > but I can't get it to work. Are you saying you've been sending us patches without ever testing them ? -- Richard Braun
Re: [PATCH] kern: simple futex for gnumach (version 3)
On 22.12.2013 18:25:10, Richard Braun wrote: > On Sun, Dec 22, 2013 at 06:22:03PM +0100, Marin Ramesa wrote: > > On 22.12.2013 17:56:32, Richard Braun wrote: > > > Test it yourself from userspace before you resubmit. > > > > That's the problem. I don't know how to test this. How to boot up a > > modified gnumach and then execute a program over it? I tried with > > QEMU, but I can't get it to work. > > Are you saying you've been sending us patches without ever > testing them ? Yes. But that's different, that's just cleaning the code. Removing forward declarations and unused variables doesn't change the behavior. Now, I need to test this, but I don't know how.
Re: [PATCH] kern: simple futex for gnumach (version 3)
On Sun, Dec 22, 2013 at 06:57:50PM +0100, Marin Ramesa wrote: > On 22.12.2013 18:25:10, Richard Braun wrote: > > Are you saying you've been sending us patches without ever > > testing them ? > > Yes. But that's different, that's just cleaning the code. Removing > forward declarations and unused variables doesn't change the behavior. > Now, I need to test this, but I don't know how. Cleaning isn't supposed to change the behaviour, but that sometimes happens, in particular with the kind of optimizations applied by a modern compiler such as gcc, so please, test before sending... -- Richard Braun
Re: [PATCH] kern: simple futex for gnumach (version 3)
On Sun, Dec 22, 2013 at 06:22:03PM +0100, Marin Ramesa wrote: > On 22.12.2013 17:56:32, Richard Braun wrote: > > > OK, I'm reading the documentation. In the meantime I have defined > > > several simple RPCs for testing purposes. I will send the updated > > > patch shortly. > > > > Test it yourself from userspace before you resubmit. > > That's the problem. I don't know how to test this. How to boot up a > modified gnumach and then execute a program over it? I tried with QEMU, > but I can't get it to work. Whether it's in a virtual machine or a real one doesn't matter at all. On Debian, simply copy the gnumach binary to /boot and run update-grub. You'll get a new entry at boot time for your kernel. -- Richard Braun
Re: [PATCH] kern: simple futex for gnumach (version 3)
On 22/12/13 22:04:15, Richard Braun wrote: > Whether it's in a virtual machine or a real one doesn't matter at > all. On Debian, simply copy the gnumach binary to /boot and run > update-grub. You'll get a new entry at boot time for your kernel. I don't have the update-grub command (I'm not using Debian). So I have to modify files by hand and when I write 'multiboot /boot/gnumach.gz' I get a 'invalid magic number' error at boot time. Can you please show me the gnumach menuentry generated by grub? Thanks.
