| Rev |
Age |
Author |
Path |
Log message |
Diff |
| 1681 |
5936d 13h |
fabio |
/shark/branches/xen/ |
Redesign the build system.
Switch to a non-recursive easy to configure build system, with a topdir
makefile that collects information on the targets to build from the subdirs
and generates the needed targets and commands. The ports and demos are
still not connected to the build. The new build system is documented
in the main makefile. |
|
| 1622 |
5966d 13h |
fabio |
/shark/branches/xen/ |
Rremove mkf/.
Noone seems to use/need it. |
|
| 1621 |
5966d 14h |
fabio |
/shark/branches/xen/ |
Move all the low level libc related headers from the OSLib to libc/.
Move the headers used to define the libc interface exported by Shark from
the OSLib to libc/arch/$(ARCH) (remember, by now ARCH can be only x86,) and
change their users to reflect the movement. There are still a couple of
problems to solve, that are how to deal with the libc code used inside the
kernel, and how to handle things that are on the boundary, e.g., some type
definitions and so on. Btw this had to be started from somewhere. |
|
| 1620 |
5966d 18h |
fabio |
/shark/branches/xen/libc/arch/x86/ |
Fix the cleanall target.
The OSLib uses an allclean target instead of cleanall, take that into
account... |
|
| 1619 |
5966d 18h |
fabio |
/shark/branches/xen/ |
Move the OSLib libc/libm code into libc/arch.
Isolate the libc/libm code inside the OSLib and move it to libc/ with
the long term goal of isolating and cleaning up the interfaces between
the system components. The ARCH configuration variable will choose
the architecture Shark is being compiled for.
Now the architecture specific part of the libc is in libc/arch/$(ARCH),
but the include/makefile support is far from being generic enough. |
|
| 1618 |
5967d 08h |
fabio |
/shark/branches/xen/ |
Started a new xen/ branch.
The plan is to support xen (without HVM) as an execution platform, and,
in the process, to decouple the generic kernel from the OSLib, trying to
make it easier to port Shark to new architectures. |
|
| 1082 |
6015d 22h |
tullio |
/shark/trunk/docs/ |
Added new changes toward version 1.5.4. |
|
| 1081 |
6015d 22h |
tullio |
/shark/trunk/config/ |
Added -fno-stack-protector flag for compiling with new gcc 4 under Ubuntu. |
|
| 1080 |
6015d 23h |
tullio |
/shark/trunk/ |
Changed the default base address to avoid changing it every time. |
|
| 1079 |
6186d 17h |
boinc |
/shark/trunk/docs/ |
Updated change log. |
|
| 1078 |
6203d 22h |
fabio |
/shark/trunk/drivers/linuxc26/ |
Fix linuxc26 timer cancellation. |
|
| 1077 |
6203d 22h |
fabio |
/shark/trunk/ |
Deactivate pending activations of interrupts. This prevents some
qemu crashes. |
|
| 1075 |
6304d 18h |
tullio |
/shark/trunk/docs/ |
Fixed typo. |
|
| 1074 |
6304d 18h |
tullio |
/shark/trunk/docs/ |
Updated with latest changes. |
|
| 1073 |
6304d 19h |
tullio |
/shark/trunk/ |
Updated makefiles to consider the new shark.cfg COMPILER option. |
|
| 1071 |
6310d 14h |
tullio |
/shark/trunk/docs/ |
Documented new changes. |
|
| 1070 |
6310d 14h |
tullio |
/shark/trunk/docs/ |
Moved from $(SHARK) to $(SHARK)/docs. |
|
| 1069 |
6310d 14h |
tullio |
/shark/trunk/ |
Moved to $(SHARK)/docs. |
|
| 1068 |
6310d 14h |
tullio |
/shark/trunk/ |
Documented changes toward rel_1_5_3. |
|
| 1067 |
6317d 14h |
tullio |
/shark/trunk/ |
Added licensing disclaimer to avoid the manual insertion of the license disclaimer in each single source file (i.e., Linux code). |
|
| 1066 |
6317d 14h |
tullio |
/shark/trunk/config/ |
Cleaned old, unused code. |
|
| 1065 |
6317d 14h |
tullio |
/shark/trunk/ports/ |
Removed due to licensing problems. |
|
| 1064 |
6317d 14h |
tullio |
/shark/trunk/ports/ |
Removed due to licensing problems. |
|
| 1063 |
6318d 22h |
tullio |
/shark/trunk/ |
Added missing GPL license disclaimer. |
|
| 1061 |
6393d 18h |
tullio |
/shark/trunk/ |
Added clearer comments to make modifications to config files easier in order to support older gcc versions. |
|
| 1060 |
6393d 19h |
tullio |
/shark/trunk/drivers/serial/ |
Fixed serial library to remove the code deadlock between init and write functions. |
|
| 1059 |
6409d 18h |
tullio |
/shark/trunk/drivers/cpu/cpufreq/ |
Removed some inlining to compile with
gcc version 4.0.2 20050901 (prerelease) (SUSE Linux) |
|
| 1057 |
6430d 03h |
tullio |
/shark/trunk/ |
Added gcc 4 support. |
|
| 1056 |
6430d 03h |
tullio |
/shark/trunk/ |
Added gcc 4x support. Actually, tested with version 4.1.2.
Into code, updated some constructs that were not compatible with gcc 4x.
Into config files, added compiler flags to suppress some warnings (turned as default since gcc 4x) and turned off optimizations for inline code, which generated unwanted calls to builtin memcpy. |
|
| 1055 |
6430d 20h |
tullio |
/shark/trunk/drivers/serial/ |
Fixed COM initialization bug about stop bit setting. |
|
| 1054 |
6486d 23h |
tullio |
/shark/trunk/ |
change log file added. |
|
| 1053 |
6487d 17h |
tullio |
/shark/trunk/ |
New roadmap agreed after the briefing in Pisa, July 2006. |
|
| 1051 |
6490d 23h |
tullio |
/shark/trunk/ |
added dynalynk support (thanks to LexN) |
|
| 1050 |
6506d 20h |
tullio |
/shark/trunk/drivers/linuxc26/include/asm/ |
Fized the multiline comment warning. |
|
| 1049 |
6507d 16h |
mauro |
/shark/trunk/drivers/usb/ |
First working version! |
|
| 1048 |
6507d 16h |
mauro |
/shark/trunk/drivers/usb/media/ |
First working vwesion! |
|
| 1047 |
6507d 17h |
mauro |
/shark/trunk/drivers/ |
Added functions needed by the USB driver. |
|
| 1046 |
6512d 16h |
tullio |
/shark/trunk/modules/intdrive/ |
The IntDrive task is now created by posting a sys_atrunlevel event at RUNLEVEL_INIT instead of creating the task directly into the INTDRIVE_register_level. |
|
| 1045 |
6515d 00h |
tullio |
/shark/trunk/kernel/ |
Fixed bug to group_activate_at which tried to activate tasks that failed during creation. Now group_activate_at tests for (proc_table[i].status != FREE) before activating tasks. |
|
| 1044 |
6528d 19h |
tullio |
/shark/trunk/drivers/linuxc26/include/asm/ |
Remap on linking traps and longer delays removed.
Should not be needed in Shark.
It eliminates unwanted delays on some machines. |
|