Quoting Samuel Thibault (2013-12-10 01:08:21)
> Justus Winter, le Mon 09 Dec 2013 15:16:35 +0100, a écrit :
> > + {
> > + pthread_mutex_lock (&(*np)->lock);
> > + netfs_nref (*np); /* Return a reference to the caller. */
> > +
Quoting Samuel Thibault (2013-12-10 01:17:02)
> Justus Winter, le Mon 09 Dec 2013 15:16:41 +0100, a écrit :
> > * console-client/trans.c (netfs_attempt_mkfile): Remove spurious
> > pthread_spin_unlock.
> > (netfs_node_norefs): Likewise.
>
> It seems completely spuriou
Previously, settrans --chroot would just exec the target. Create a
new process for that purpose. Wait for its completion, then ask the
translator (nicely by default) to go away. If it refuses with EBUSY,
it might be because some process has daemonized inside the chrooted
environment.
This fixes
Previously, files created in the fakeroot environment were created
with the uid and gid of the user running fakeroot:
% fakeroot-hurd /bin/sh -c 'touch /tmp/$$; stat --format=%u:%g /tmp/$$'
1000:1000
* trans/fakeroot.c (netfs_S_dir_lookup): Fix ownership of newly
created files.
---
trans/fakeroo
* trans/fakeroot.c (netfs_S_dir_lookup): Drop else.
---
trans/fakeroot.c | 76 ++
1 file changed, 37 insertions(+), 39 deletions(-)
diff --git a/trans/fakeroot.c b/trans/fakeroot.c
index 6882618..f278db1 100644
--- a/trans/fakeroot.c
+++ b/tra
* trans/fakeroot.c (netfs_S_dir_lookup): Remove dead code.
---
trans/fakeroot.c | 67 +++---
1 file changed, 29 insertions(+), 38 deletions(-)
diff --git a/trans/fakeroot.c b/trans/fakeroot.c
index f278db1..58f1112 100644
--- a/trans/fakeroot.c
++
* trans/fakeroot.c (netfs_S_dir_lookup): Remove dead code.
---
trans/fakeroot.c |6 ++
1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/trans/fakeroot.c b/trans/fakeroot.c
index 58f1112..203f2c1 100644
--- a/trans/fakeroot.c
+++ b/trans/fakeroot.c
@@ -371,10 +371,8 @@ netfs_S_
Previously, settrans --chroot would just exec the target. Create a
new process for that purpose. Wait for its completion, then ask the
translator (nicely by default) to go away. If it refuses with EBUSY,
it might be because some process has daemonized inside the chrooted
environment.
This fixes
err, rather this one, which honors goaway_flags.
Justus
ping ;)
Justus
* utils/rpctrace.c (rewrite_right): Generalize the code so we can use
rewrite_right to rewrite MACH_MSG_TYPE_PORT_SEND rights for non-rpc ports.
---
utils/rpctrace.c |4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/utils/rpctrace.c b/utils/rpctrace.c
index ef7ab53..9433ef3
Currently, rpctrace dies if a tracee uses select(2) because it asserts
that reply_type is a MACH_MSG_TYPE_PORT_SEND_ONCE right. Generalize
the code surrounding the failing assertion.
* utils/rpctrace.c (trace_and_forward): Generalize code by switching
over reply_type.
---
utils/rpctrace.c | 40
This allows one to rpctrace processes doing select(2).
* utils/rpctrace.c (trace_and_forward): Handle MACH_MSG_TYPE_PORT_SEND
rights.
---
utils/rpctrace.c |4
1 file changed, 4 insertions(+)
diff --git a/utils/rpctrace.c b/utils/rpctrace.c
index 5d3b2ca..b39b2e3 100644
--- a/utils/rpctr
mount uses libblkid (if available) to detect the file system type when
using mount -t auto. libblkid calls fat file systems "vfat", our fat
translator is called "fatfs". Fix this discrepancy in mount.
* utils/mount.c (do_mount): Fix autodetection of fat file systems.
---
utils/mount.c |7 ++
* utils/rpctrace.c (escape_sequences): New char array mapping
characters to their escape sequence.
(print_data): Escape non-printable characters when printing strings.
---
utils/rpctrace.c | 48 ++--
1 file changed, 46 insertions(+), 2 deletions(-)
di
Currently, it is impossible to properly attribute response messages to
requests. Even though rpctrace is single-threaded, its tracee may
not. Or there might be more than one tracee. In any such case it is
not guaranteed that the reply message we just processed is for the
request we just printed.
Quoting Ivan Shmakov (2013-12-13 17:05:19)
> >>>>> Justus Winter <4win...@informatik.uni-hamburg.de> writes:
>
> > * utils/rpctrace.c (escape_sequences): New char array mapping
> > characters to their escape sequence.
>
> […]
>
> > +
* xen/console.h: Add missing includes.
---
xen/console.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/xen/console.h b/xen/console.h
index 2b78f29..061ba28 100644
--- a/xen/console.h
+++ b/xen/console.h
@@ -21,7 +21,9 @@
#include
#include
+#include
#include
+#include
#define h
Hi :)
I've been playing around with buildbot, the goal is to automatically
produce Debian packages from the upstream git repositories. This is
work in progress.
To that end I took the Debian repositories of Hurd, MIG and GNU Mach
and removed anything that is present in the upstream repositories.
* i386/i386at/kd_event.h: Add missing includes.
---
i386/i386at/kd_event.h | 4
1 file changed, 4 insertions(+)
diff --git a/i386/i386at/kd_event.h b/i386/i386at/kd_event.h
index f1295cb..9568fa5 100644
--- a/i386/i386at/kd_event.h
+++ b/i386/i386at/kd_event.h
@@ -26,6 +26,10 @@
#ifndef _KD
Quoting Justus Winter (2013-12-18 12:38:31)
> While doing that I discovered that the recent cleanups broke GNU Mach
> when compiled for the Xen target. Two issues were merely missing
> includes, but now it fails like this:
>
> % ./configure --enable-kdb --enable-platform=xen --enab
Previously, fakeroot would not exit if a process outlived the original
process started by settrans. This caused bugs like this:
% fakeroot-hurd /bin/sh -c 'sleep 1&' 2>&1 | tee
Fix this by exiting if the last client of fakeroot goes away.
* trans/fakeroot.c (fakeroot_netfs_release_protid): Shu
Previously, fakeroot would not exit if a process outlived the original
process started by settrans. This caused bugs like this:
% fakeroot-hurd /bin/sh -c 'sleep 1&' 2>&1 | tee
Fix this by exiting if the last client of fakeroot goes away.
If noone has a right to any control or protid port, noo
Richard mentioned a use case I had not considered, binding
/hurd/fakeroot to a node rather than using settrans --chroot. The
patch now covers this use case by also checking for control ports.
Justus
Apparently, the NS32000 was a 32-bit CPU from the 1990ies. The string
"ns32000" appears nowhere else in the source.
* vm/vm_page.h (struct vm_page): Remove NS32000-specific padding.
---
vm/vm_page.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/vm/vm_page.h b/vm/vm_page.h
index 4445cb0..
* vm/vm_page.h (struct vm_page): Merge the two bit fields.
---
vm/vm_page.h | 10 +++---
1 file changed, 3 insertions(+), 7 deletions(-)
diff --git a/vm/vm_page.h b/vm/vm_page.h
index 7392044..339d355 100644
--- a/vm/vm_page.h
+++ b/vm/vm_page.h
@@ -92,11 +92,8 @@ struct vm_page {
Previously, the bit field left 31 bits unused. By reducing the size
of wire_count by one bit, the size of the whole struct is reduced by
four bytes.
* vm/vm_page.h (struct vm_page): Reduce the size of wire_count to 15
bits.
---
vm/vm_page.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Previously, the size of struct kmem_cache was 136 bytes, just eight
bytes larger than 128 bytes, which is typically two cache lines on
today's CPUs.
By reducing the size of the name field which holds a human-readable
description by eight bytes to 24 bytes, the struct kmem_cache can be
made fit int
* kern/task.h (struct task): Reduce the size of struct task by
2 * sizeof boolean_t by using a bit field for the boolean flags.
---
kern/task.h | 8 +---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/kern/task.h b/kern/task.h
index 9bfea57..e852033 100644
--- a/kern/task.h
+++
Hi :)
Quoting Marin Ramesa (2014-01-04 11:12:42)
> This is not yet ready to be called from userspace since glibc
> build on the Hurd doesn't seem to compile the new RPCs. I don't
> know how to actually call them.
mig is your friend:
% mig pong.defs
% ls pong*
pong.defs pong.h pongServer.c pon
The pattern is !x&y. An expression of this form is almost always
meaningless, because it combines a boolean operator with a bit
operator. In particular, if the rightmost bit of y is 0, the result
will always be 0.
Fixed using coccinelle.
// !x&y combines boolean negation with bitwise and
//
// Co
* kern/slab.h (KMEM_CACHE_NAME_SIZE): Explain the significance of the
chosen length.
---
kern/slab.h | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/kern/slab.h b/kern/slab.h
index a4219c4..fd65893 100644
--- a/kern/slab.h
+++ b/kern/slab.h
@@ -147,7 +147,9 @@ typedef vm_off
* kern/slab.h (struct kmem_cache): Reorder the fields so that all hot
fields are within the first cache line.
---
kern/slab.h | 18 --
1 file changed, 12 insertions(+), 6 deletions(-)
diff --git a/kern/slab.h b/kern/slab.h
index fd65893..457dee4 100644
--- a/kern/slab.h
+++ b/kern
The pattern is !x&y. An expression of this form is almost always
meaningless, because it combines a boolean operator with a bit
operator. In particular, if the rightmost bit of y is 0, the result
will always be 0.
Fixed using coccinelle.
// !x&y combines boolean negation with bitwise and
//
// Co
Quoting Samuel Thibault (2014-01-06 02:43:20)
> Justus Winter, le Mon 06 Jan 2014 00:34:55 +0100, a écrit :
> > * kern/slab.h (struct kmem_cache): Reorder the fields so that all hot
> > fields are within the first cache line.
>
> Did you try to benchmark this a bit?
>
Quoting Samuel Thibault (2014-01-06 02:41:03)
> > * linux/src/drivers/net/tlan.c: Fix bit tests.
> > * linux/src/drivers/scsi/AM53C974.c: Likewise.
> > * linux/src/drivers/scsi/FlashPoint.c: Likewise.
> > * linux/src/drivers/scsi/NCR5380.c: Likewise.
> > * linux/src/drivers/scsi/t128.c: Likewise.
>
* include/cache.h (__cacheline_aligned): This macro can be used to
align statically allocated objects so that they start at a cache line.
---
include/cache.h | 25 +
1 file changed, 25 insertions(+)
create mode 100644 include/cache.h
diff --git a/include/cache.h b/include
* device/dev_lookup.c: Align kmem_cache objects using __cacheline_aligned.
* device/dev_pager.c: Likewise.
* device/ds_routines.c: Likewise.
* device/net_io.c: Likewise.
* i386/i386/fpu.c: Likewise.
* i386/i386/machine_task.c: Likewise.
* i386/i386/pcb.c: Likewise.
* i386/intel/pmap.c: Likewise.
*
Previously the mtab detector tried to detect whether it talks to
itself by comparing the control port of the translator it currently
visits with its own. This was done to prevent deadlocks. However, this
did not detect non-trivial loops, i.e. two mtab translators would
still dead-lock each other.
..9b867b7 100644
--- a/trans/mtab.c
+++ b/trans/mtab.c
@@ -1,6 +1,6 @@
/* This is an mtab translator.
- Copyright (C) 2013 Free Software Foundation, Inc.
+ Copyright (C) 2013,14 Free Software Foundation, Inc.
Written by Justus Winter <4win...@informatik.uni-hamburg.de>
@@ -594,7 +
* trans/mtab.c (main): Use ports_manage_port_operations_multithread.
---
trans/mtab.c | 10 +++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/trans/mtab.c b/trans/mtab.c
index 5a5a07d..adfb345 100644
--- a/trans/mtab.c
+++ b/trans/mtab.c
@@ -235,9 +235,13 @@ main (int argc,
* trans/mtab.c (struct mtab): Add lock.
(open_hook): Initialize lock.
(close_hook): Destroy lock.
(trivfs_S_io_read): Lock mtab object, adjust error handling accordingly.
(trivfs_S_io_seek): Likewise.
(trivfs_S_io_readable): Likewise.
---
trans/mtab.c | 44 -
Traditionally, /proc/mounts includes only filesystem mount points.
Previously though, the mtab translator included any kind of
translator, like all /hurd/storeio translators. This causes problems
with umount --all as this would remove the passive translator records
from nodes in /dev.
Fix this by
Detect passive translator startup and add the resulting active
translator to the list of active translators.
Note that newpi is properly deallocated in the function epilogue.
* libdiskfs/dir-lookup.c (diskfs_S_dir_lookup): Detect and register
passive translator startup.
---
libdiskfs/dir-looku
Detect passive translator startup and add the resulting active
translator to the list of active translators.
* libnetfs/dir-lookup.c (netfs_S_dir_lookup): Detect and register
passive translator startup.
---
libnetfs/dir-lookup.c | 43 +--
1 file changed,
fshelp/translator-list.c
index 87dcb21..3ece711 100644
--- a/libfshelp/translator-list.c
+++ b/libfshelp/translator-list.c
@@ -1,6 +1,6 @@
/* A list of active translators.
- Copyright (C) 2013 Free Software Foundation, Inc.
+ Copyright (C) 2013,14 Free Software Foundation, Inc.
Hi Ignazio :)
Quoting Ignazio Sgalmuzzo (2014-01-12 23:22:37)
> While testing, I noticed a little problem with "settrans -ca test
> /hurd/storeio -T bunzip2 test.bz2": it seems it looks for a .so module
> with wrong name (something like "libstore_test.bz2.so.0.3)": I want to
> investigate furth
Hi Marin :)
Quoting Marin Ramesa (2014-01-13 15:38:19)
> Fixed a bug in private wakeups by calling ipc_entry_alloc_name().
>
> This is the last version as everything works now.
Heh, how many times did I think just that >,<
> To do a quick userspace mutex test with timed waits:
Good to see you
Quoting Marin Ramesa (2014-01-13 19:15:41)
> Thanks for your reviews, I learned a lot.
You are welcome :)
> I won't take up your time anymore.
Why not? We are all consenting adults here, right? Anyone can decide
by herself how much time she wants to spend on stuff.
> I will do this privately
Quoting Marin Ramesa (2014-01-15 01:48:01)
> So, I decided to keep working on this.
Awesome :)
> [..]
> I tested a simple userspace mutex based on this and multiple futexes,
> all tests passed.
Could you put that somewhere? I still have so little "context", I'd
like to see how futexe are used i
Quoting Marin Ramesa (2014-01-15 14:12:08)
> On 01/15/2014 10:56:25 AM, Justus Winter wrote:
> > > I tested a simple userspace mutex based on this and multiple
> > futexes,
> > > all tests passed.
> >
> > Could you put that somewhere? I still have so li
Quoting Justus Winter (2013-12-02 17:25:23)
> Quoting Samuel Thibault (2013-12-02 13:09:44)
> > Well, it means that whenever a protocol gets augmented, one has to
> > recompile all demuxer functions calling these inlines. In the example
> > of trivfs_demuxer, the application
It is not clear to me why the declaration was put there in the first
place. It is not used anywhere, and it conflicts with the declaration
generated by mig.
* vm/memory_object_proxy.h (memory_object_create_proxy): Remove declaration.
---
vm/memory_object_proxy.h | 11 ---
1 file changed,
GNU MIG recently gained support for emitting x_server_routine
declarations in the generated server header file. Using this
declaration, the x_server_routine functions can be inlined into the
ipc_kobject_server function.
* kern/ipc_kobject.c: Include the mig-generated server headers.
---
kern/ipc
Quoting Marin Ramesa (2014-01-17 15:32:05)
> Function __atomic_load_n() is now used. Private futex structure and queue
> was defined to avoid conflicts in block and wakeup events. Now, private
> futex is really private and may only be used by threads of the task where
> it was locked.
>
> All of l
Use translation functions instead of doing the lookup manually.
* trans/Makefile (password-MIGSFLAGS): Add mutators.
* trans/password.c (S_password_check_user): Update accordingly.
(S_password_check_group): Likewise.
---
trans/Makefile | 6 +-
trans/password.c | 30 ++--
dex 000..0d5bd4d
--- /dev/null
+++ b/proc/mig-decls.h
@@ -0,0 +1,42 @@
+/* Translation functions for mig.
+
+ Copyright (C) 2013 Free Software Foundation, Inc.
+
+ Written by Justus Winter <4win...@informatik.uni-hamburg.de>
+
+ This file is part of the GNU Hurd.
+
+ The GNU Hurd
GNU MIG recently gained support for emitting x_server_routine
declarations in the generated server header file. Using this
declaration, the x_server_routine functions can be inlined into the
demuxer function.
* auth/auth.c: Include the mig-generated server headers.
---
auth/auth.c | 10 -
GNU MIG recently gained support for emitting x_server_routine
declarations in the generated server header file. Using this
declaration, the x_server_routine functions can be inlined into the
demuxer function.
* utils/fakeauth.c: Include the mig-generated server headers.
---
utils/fakeauth.c | 6
The canonical name for variables of the type error_t is err. There
are, however, places where the variable is called error instead. This
is unfortunate, as this shadows the error function. Rename such
variables to err.
For reference, this is accomplished using the following semantic
patch:
@@
Quoting Samuel Thibault (2014-01-19 21:14:03)
> Justus Winter, le Sat 18 Jan 2014 18:19:26 +0100, a écrit :
> > * proc/mgt.c (S_proc_exception_raise): Update accordingly.
> > * proc/mig-decls.h: New file.
> > * proc/proc_exc.defs: Add translation functions.
>
> Why pu
Previously, a banner was hardcoded in Hurds getty. Load /etc/issue
instead. Fall back to a hardcoded version if that fails. Expand
common \char sequences. If no /etc/issue is provided, the old
behavior is retained.
* daemons/getty.c (load_banner): New function.
(print_banner): Expand \char seq
* kern/printf.c (snprintf): New function.
* kern/printf.h (snprintf): New declaration.
---
kern/printf.c | 10 ++
kern/printf.h | 1 +
2 files changed, 11 insertions(+)
diff --git a/kern/printf.c b/kern/printf.c
index af59d5a..ea78d48 100644
--- a/kern/printf.c
+++ b/kern/printf.c
@@ -61
The name field is not necessarily zero-terminated. Use a precision
argument to limit the number of characters written when printing the
name.
* kern/slab.c (kmem_cache_error): Fix printing of cache names.
---
kern/slab.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/kern/
* ipc/mach_port.c (mach_port_destroy): Use the name of the task for
error messages.
(mach_port_deallocate): Likewise.
---
ipc/mach_port.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/ipc/mach_port.c b/ipc/mach_port.c
index fbc5e69..13572a1 100644
--- a/ipc/mach_port.c
++
* kern/task.c (task_create): Initialize name with the address of the task.
* kern/task.h (TASK_NAME_SIZE): New definition.
(struct task): Add field name.
---
kern/task.c | 3 +++
kern/task.h | 9 +
2 files changed, 12 insertions(+)
diff --git a/kern/task.c b/kern/task.c
index 13b3c76..0b5
* include/mach/debug.defs: New file.
* include/mach/debug.h: Likewise.
---
include/mach_debug/mach_debug_types.defs | 2 ++
include/mach_debug/mach_debug_types.h| 11 +++
2 files changed, 13 insertions(+)
diff --git a/include/mach_debug/mach_debug_types.defs
b/include/mach_debug/mac
task_set_name sets the name of a task. This is a debugging aid. The
name will be used in error messages printed by the kernel.
* kern/task.c (task_set_name): New function.
* kern/task.h (task_set_name): New declaration.
---
kern/task.c | 16
kern/task.h | 4
2 files chang
task_set_name sets the name of a task. This is a debugging aid. The
name will be used in error messages printed by the kernel.
* include/mach/gnumach.defs (task_set_name): New procedure.
---
include/mach/gnumach.defs | 9 +
1 file changed, 9 insertions(+)
diff --git a/include/mach/gnum
* doc/mach.texi (Task Information): Document the new task_set_name
procedure.
---
doc/mach.texi | 11 +++
1 file changed, 11 insertions(+)
diff --git a/doc/mach.texi b/doc/mach.texi
index 9ad9e70..d089224 100644
--- a/doc/mach.texi
+++ b/doc/mach.texi
@@ -5037,6 +5037,17 @@ total system r
Quoting Samuel Thibault (2014-02-01 15:20:55)
> Justus Winter, le Sat 01 Feb 2014 15:09:21 +0100, a écrit :
> > The name field is not necessarily zero-terminated. Use a precision
> > argument to limit the number of characters written when printing the
> > name.
>
> M
Quoting Samuel Thibault (2014-02-02 00:11:59)
> Justus Winter, le Sat 01 Feb 2014 16:34:20 +0100, a écrit :
> > Quoting Samuel Thibault (2014-02-01 15:20:55)
> > > Justus Winter, le Sat 01 Feb 2014 15:09:21 +0100, a écrit :
> > > > The name field is not neces
Hi :)
Quoting Thomas Schwinge (2013-10-25 15:48:06)
> When I recently read about it somewhere, I've also had the idea about
> feeding the Hurd code into the Coverity scanner, which I think offers
> such a service for Free Software projects. I also thought about dping
> the same for GNU Mach and g
task_set_name sets the name of a task. This is a debugging aid. The
name will be used in error messages printed by the kernel.
* kern/task.c (task_set_name): New function.
* kern/task.h (task_set_name): New declaration.
---
kern/task.c | 17 +
kern/task.h | 4
2 files chan
* kern/slab.c (kmem_cache_error): Use kmem_warn instead of kmem_error
to print the cache name and its address.
---
kern/slab.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kern/slab.c b/kern/slab.c
index d1e3632..adf07e7 100644
--- a/kern/slab.c
+++ b/kern/slab.c
@@ -662,7 +
Use strncpy instead of strcpy to copy the name of a symbol into the
symtab structure. Make sure that the string is properly terminated.
Found using Coverity.
* ddb/db_sym.c (db_add_symbol_table): Use strncpy instead of strcpy,
ensure string termination.
---
ddb/db_sym.c | 3 ++-
1 file changed,
The slab allocator relies on the fact that kmem_cache_error does not
return. Previously, kmem_error was using printf. Use panic instead.
Found using the Clang Static Analyzer.
* kern/slab.c (kmem_error): Use panic instead of printf.
---
kern/slab.c | 4 ++--
1 file changed, 2 insertions(+), 2
Previously, only strlen(device_name) bytes were allocated, missing one
byte for the terminating zero.
* xen/block.c (hyp_block_init): Fix buffer size.
---
xen/block.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/xen/block.c b/xen/block.c
index 3e4ce7c..bd3758f 100644
--- a/
* exec/main.c (S_exec_init): Get the device master port and connect
std{in,out,err} to the mach console.
---
exec/main.c | 17 +++--
1 file changed, 15 insertions(+), 2 deletions(-)
diff --git a/exec/main.c b/exec/main.c
index c1f347c..27f33b1 100644
--- a/exec/main.c
+++ b/exec/main.
* daemons/getty.c (load_banner): Drop first newline from default banner.
(print_banner): Print a newline.
---
daemons/getty.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/daemons/getty.c b/daemons/getty.c
index 7112660..40ad4d7 100644
--- a/daemons/getty.c
+++ b/daemons/ge
Quoting Samuel Thibault (2014-02-05 01:25:40)
> Justus Winter, le Wed 05 Feb 2014 01:07:46 +0100, a écrit :
> > * exec/main.c (S_exec_init): Get the device master port and connect
> > std{in,out,err} to the mach console.
>
> Mmm, stdin too, really?
I haven't thought a
Quoting Samuel Thibault (2014-02-05 18:21:48)
> Ludovic Courtès, le Wed 05 Feb 2014 14:52:27 +0100, a écrit :
> > Justus Winter <4win...@informatik.uni-hamburg.de> skribis:
> >
> > > * exec/main.c (S_exec_init): Get the device master port and connect
> > &
Hi Ignazio :)
Quoting Samuel Thibault (2014-02-09 21:39:07)
> Hello,
>
> Ignazio Sgalmuzzo, le Sat 04 Jan 2014 22:59:25 +, a écrit :
> > configure.ac: added check for libbz2
> > libstore/Makefile: linked libbz2
> > ext2fs/Makefile: linked libbz2
> > do-bunzip2.c: rewritten do_bunzip2 using li
* isofs/Makefile (OTHERLIBS): Add bz2.
---
isofs/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/isofs/Makefile b/isofs/Makefile
index 053b622..a037c21 100644
--- a/isofs/Makefile
+++ b/isofs/Makefile
@@ -22,7 +22,7 @@ SRCS = inode.c main.c lookup.c pager.c rr.c
OBJ
* fatfs/Makefile (OTHERLIBS): Add bz2.
---
fatfs/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fatfs/Makefile b/fatfs/Makefile
index c673b1b..9a04fbb 100644
--- a/fatfs/Makefile
+++ b/fatfs/Makefile
@@ -23,7 +23,7 @@ SRCS = inode.c main.c dir.c pager.c fat.c virt-ino
Hi :)
when I implemented the mtab stuff last summer I made quite a blunder.
I added fsys_get_{children,source}, but send those messages to ports
associated with nodes, not to control ports. So:
[PATCH 1/6] hurd: fix the get-children and get-source procedures
Removes fsys_get_{children,source}, a
---
trans/Makefile | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/trans/Makefile b/trans/Makefile
index c0386d0..90df479 100644
--- a/trans/Makefile
+++ b/trans/Makefile
@@ -56,7 +56,7 @@ magic: ../libiohelp/libiohelp.a
hello: ../libtrivfs/libtrivfs.a ../libfshelp/libfshelp.a
Previously, map_device_to_path mapped unknown device strings to
"none". Return a copy of the original source string instead so that
e.g. nfs mount point points are properly handled.
* trans/mtab.c (map_device_to_path): Do not map unknown source strings
to "none".
---
trans/mtab.c | 2 +-
1 file
Implement netfs_get_source so that nfs translators show up in
/proc/mounts.
* nfs/main.c (netfs_get_source): New function.
---
nfs/main.c | 14 ++
1 file changed, 14 insertions(+)
diff --git a/nfs/main.c b/nfs/main.c
index 3ce75a9..cd1c29a 100644
--- a/nfs/main.c
+++ b/nfs/main.c
@@
Previously, the mtab translator would guess the source of a filesystem
translator from its last argument. This made all kinds of
non-filesystem translators appear in /proc/mounts, causing severe
problems due to umount --all removing vital passive translator
records.
Fix this by ignoring all trans
By default diskfs_disk_name is returned as the source of
libdiskfs-based translators. Hurds fstab allows multiple entries with
the mnt_fsname being "none".
* tmpfs/tmpfs.c (diskfs_disk_name): Set to "none".
---
tmpfs/tmpfs.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tmp
@@ -0,0 +1,95 @@
+/* file_get_children
+
+ Copyright (C) 2013 Free Software Foundation, Inc.
+
+ Written by Justus Winter <4win...@informatik.uni-hamburg.de>
+
+ This file is part of the GNU Hurd.
+
+ The GNU Hurd is free software; you can redistribute it and/or modify
+ it under the
* libdiskfs/diskfs.h (struct diskfs_control): New declaration.
(diskfs_begin_using_control_port): New declaration and function.
(diskfs_end_using_control_port): Likewise.
* libdiskfs/fsmutations.h: Add translation functions.
* libdiskfs/priv.h (control_t): New type declaration for mig.
* libdiskfs/
* libdiskfs/boot-start.c (diskfs_S_fsys_init): Fix typo.
---
libdiskfs/boot-start.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/libdiskfs/boot-start.c b/libdiskfs/boot-start.c
index 6d9a773..05d3b43 100644
--- a/libdiskfs/boot-start.c
+++ b/libdiskfs/boot-start.c
@@ -460,7
* mutations.h: Add translation functions.
* netfs.h (struct netfs_control): New declaration.
* priv.h: Define translation functions.
* fsys-get-options.c: Fix receiver lookups.
* fsys-getroot.c: Likewise.
* fsys-goaway.c: Likewise.
* fsys-set-options.c: Likewise.
* fsys-syncfs.c: Likewise.
* fsysst
Hi :)
Quoting Samuel Thibault (2014-02-12 18:27:21)
> Could you mention the rationale of these changes? It's far from obvious
> for anybody who does not know RPCs very well :)
This is a prelude to my protected payload patch series. I'm fixing
the receiver lookups that are currently done "by foo
Hi :)
this patch series implements protected payloads on top of the mach
messaging system. It preserves the semantics for programs not using
this functionality.
This is the kernel part of my work that will (hopefully) speed up the
Hurd. For a detailed sketch see:
id:20131128235647.1092.90...@t
* include/mach/mach_port.defs: Add mach_port_set_protected_payload.
* ipc/mach_port.c: Implement mach_port_set_protected_payload.
---
include/mach/mach_port.defs | 9 +
ipc/mach_port.c | 37 +
2 files changed, 46 insertions(+)
diff --git a/
* doc/mach.texi (Receive Rights): Document
mach_port_set_protected_payload.
---
doc/mach.texi | 18 ++
1 file changed, 18 insertions(+)
diff --git a/doc/mach.texi b/doc/mach.texi
index d089224..67c5fe9 100644
--- a/doc/mach.texi
+++ b/doc/mach.texi
@@ -2715,6 +2715,24 @@ In additi
* include/mach/mach_port.defs: Skip the routines mach_port_set_rpcinfo
and mach_port_create_act if MIGRATING_THREADS is not defined.
---
include/mach/mach_port.defs | 5 +
1 file changed, 5 insertions(+)
diff --git a/include/mach/mach_port.defs b/include/mach/mach_port.defs
index e1f45e3..7
501 - 600 of 1489 matches
Mail list logo
