Commit graph

3313 commits

Author SHA1 Message Date
Thomas Veerman
4b6b1bc47a Fix typo 2010-06-07 13:46:05 +00:00
Ben Gras
a98b9ca8dc tools: fix installing includes. 2010-06-07 10:10:27 +00:00
Tomas Hruby
cbc9586c13 Lazy FPU
- FPU context is stored only if conflict between 2 FPU users or while
  exporting context of a process to userspace while it is the active
  user of FPU

- FPU has its owner (fpu_owner) which points to the process whose
  state is currently loaded in FPU

- the FPU exception is only turned on when scheduling a process which
  is not the owner of FPU

- FPU state is restored for the process that generated the FPU
  exception. This process runs immediately without letting scheduler
  to pick a new process to resolve the FPU conflict asap, to minimize
  the FPU thrashing and FPU exception hadler execution

- faster all non-FPU-exception kernel entries as FPU state is not
  checked nor saved

- removed MF_USED_FPU flag, only MF_FPU_INITIALIZED remains to signal
  that a process has used FPU in the past
2010-06-07 07:43:17 +00:00
Arun Thomas
b641afc78a VM: Remove legacy non-paging code paths 2010-06-05 14:39:40 +00:00
Erik van der Kouwe
e7e6508854 Enter package names in packman 2010-06-05 13:25:41 +00:00
Cristiano Giuffrida
354d88f883 Put initialization code where it belongs. 2010-06-04 18:08:15 +00:00
Cristiano Giuffrida
a53514d4a9 Fix range checking in safecopy. 2010-06-04 18:05:38 +00:00
Arun Thomas
8c69c6cd7f Remove gcc-4.1.1 from packages list 2010-06-04 12:36:40 +00:00
Tomas Hruby
f28acecb78 Removed a buggy assert unintentionally commted in r7044 2010-06-04 10:54:43 +00:00
Erik van der Kouwe
8afc228c48 This patch changes the release script in the following ways:
- A staging directory is always used to avoid oversized images;
- As a consequence, the zero-filling is removed so no more "out of 
  space" errors should be printed to the console;
- The root and usr partition sizes are computed so less space should be 
  wasted (the root partition gets extra 1MB zones and 64 inodes for 
  run-time though and hardlinks/holes make the used space slightly less
  than expected); USRMB (and the new ROOTMB) are now used to enforce 
  a minimum size rather than set the size;
- TMPDISK1-3 are renamed to more meaningful names (and TMPDISK2 is 
  dropped because a separate tmp directory is no longer needed);
- The ramdisks are truncated at the end to save memory (not sure 
  whether it is actually released though).
2010-06-04 05:09:44 +00:00
Erik van der Kouwe
1bb7c4d78a Create link for lstat 2010-06-04 04:44:09 +00:00
Ben Gras
2f892aca91 kernel fpu context switching: fix race condition
There seems to have been a broken assumption in the fpu context
restoring code.  It restores the context of the running process, without
guarantee that the current process is the one that will be scheduled.
This caused fpu saving for a different process to be triggered without
fpu hardware being enabled, causing an fpu exception in the kernel. This
practically only shows up with DEBUG_RACE on. Fix my thruby+me.

The fix
 . is to only set the fpu-in-use-by-this-process flag in the
   exception handler, and then take care of fpu restoring when
   actually returning to userspace

And the patch
 . translates fpu saving and restoring to c in arch_system.c,
   getting rid of a juicy chunk of assembly
 . makes osfxsr_feature private to arch_system.c
 . removes most of the arch dependent code from do_sigsend
2010-06-03 11:32:22 +00:00
Cristiano Giuffrida
332842295a Always skip signals for PM itself when broadcasting. 2010-06-03 11:18:43 +00:00
Ben Gras
035c5a369c move drivers/random SHA implementation to libutil and header to <minix/sha2.h> 2010-06-02 22:58:29 +00:00
Kees van Reeuwijk
36e12d5bd8 Use endpoint_t for the destination of mini_send and _syscall, and the
source of mini_receive.

Also some small cleanup.
2010-06-02 21:51:32 +00:00
Gianluca Guida
0a42d36ac6 Fix ftp client's passive mode.
Before the ioctl the code zeroed the wrong variable.
2010-06-02 21:12:54 +00:00
Kees van Reeuwijk
ed0b81c25c Removed some unused variables and functions. 2010-06-02 19:41:38 +00:00
Erik van der Kouwe
fc29251444 setup: default root partition size is 64MB, reinstall keeps old root size 2010-06-02 17:07:10 +00:00
Tomas Hruby
463be393c2 Gracefull crash if FPU exception in kernel 2010-06-02 13:59:55 +00:00
Ben Gras
495970a17c boot monitor: introduce a 'reset' command that causes the cpu to
reset (triple fault).
2010-06-02 13:23:15 +00:00
Tomas Hruby
7bfa47476c Fixed stack trace if panic in stop_context() 2010-06-02 08:53:49 +00:00
Arun Thomas
ec27911a7f Ramdisk makefile update 2010-06-01 15:11:43 +00:00
Arun Thomas
2fd1f18803 Release script: Remove .d/.depend files from CD 2010-06-01 14:44:36 +00:00
Thomas Veerman
6bbcab3ec4 Clean up MFS a bit:
- Remove unused includes.
 - Add include guards to headers.
 - Use unsigned variables in case they're never going to hold a negative
   value. This causes GCC's complaints to disappear and should make flexelint
   a lot happier, too.
 - Make functions private when they're used only within a module.
 - Remove unused variables.
 - Add casts where appropriate.
2010-06-01 12:35:33 +00:00
Ben Gras
cc6fed4c51 packages: don't put oss on cd. 2010-06-01 11:11:35 +00:00
Arun Thomas
8579c59b3f Minor makefile changes for GCC 2010-06-01 09:41:31 +00:00
Tomas Hruby
40f440b8cd KCall methods do not depend on m_source and m_type fields
- substituted the use of the m_source message field by
  caller->p_endpoint in kernel calls. It is the same information, just
  passed more intuitively.
  
- the last dependency on m_type field is removed.
  
- do_unused() is substituted by a check for NULL.

- this pretty much removes the depency of kernel calls on the general
  message format. In the future this may be used to pass the kcall
  arguments in a different structure or registers (x86-64, ARM?) The
  kcall number may be passed in a register already.
2010-06-01 08:54:31 +00:00
Tomas Hruby
ebbd319ac0 do_safecopy split
- removes dependency of do_safecopy() on the m_type field of the kcall
  messages.

- instead of do_safecopy() figuring out what action is requested, the
  correct safecopy method is called right away.
2010-06-01 08:51:37 +00:00
Erik van der Kouwe
8bca982581 Remove those annoying $Id and $Revision tags 2010-05-31 20:00:08 +00:00
Thomas Veerman
5d78cefaf2 Clean up PFS 2010-05-28 09:39:18 +00:00
Erik van der Kouwe
8d06457241 Remove obsolete _fpu_present variable from crtso.S 2010-05-27 09:46:42 +00:00
Erik van der Kouwe
a31e9b0400 Oops, committed too much 2010-05-27 09:20:50 +00:00
Erik van der Kouwe
fc7438181b Avoid use of C++ reserved word class in headers (reported by Aki Goto, tracker item 375) 2010-05-27 09:18:49 +00:00
Erik van der Kouwe
43b589c1cc Avoid use of C++ reserved word class in headers (reported by Aki Goto, tracker item 375) 2010-05-27 08:48:53 +00:00
Arun Thomas
b48b037fbe Reorganize some directories
servers/hgfs/hgfs_server => servers/hgfs
servers/hgfs/libhgfs => lib/libhgfs
servers/rs/service => commands/service
drivers/memory/memory_driver => drivers/memory
drivers/memory/ramdisk => drivers/ramdisk
2010-05-26 22:49:57 +00:00
Arun Thomas
007104d60e GCC build fixes/updates
-Set stack sizes for boot image processes
-Increase RS stack size
-Reduce ramdisk size
-HARDWARE task should use kernel stack
-Minor asm tweaks for leading underscores
2010-05-26 18:45:55 +00:00
David van Moolenbroek
b16fc7e7bf different rtl8169 shutdown code 2010-05-26 11:22:06 +00:00
David van Moolenbroek
7f98ba962a make IS report masked IRQs properly 2010-05-26 08:44:50 +00:00
Tomas Hruby
24764ff47a Fixed ms-based scheduling for legacy timer 2010-05-26 08:20:29 +00:00
Tomas Hruby
a8111c5027 Various small scheduling related fixes 2010-05-26 07:16:39 +00:00
David van Moolenbroek
51ff10d7c0 reset alarm timer on PRIVCTL 2010-05-26 07:10:28 +00:00
Tomas Hruby
451a6890d6 scheduling - time quantum in miliseconds
- Currently the cpu time quantum is timer-ticks based. Thus the
  remaining quantum is decreased only if the processes is interrupted
  by a timer tick. As processes block a lot this typically does not
  happen for normal user processes. Also the quantum depends on the
  frequency of the timer.

- This change makes the quantum miliseconds based. Internally the
  miliseconds are translated into cpu cycles. Everytime userspace
  execution is interrupted by kernel the cycles just consumed by the
  current process are deducted from the remaining quantum.

- It makes the quantum system timer frequency independent.

- The boot processes quantum is loosely derived from the tick-based
  quantas and 60Hz timer and subject to future change

- the 64bit arithmetics is a little ugly, will be changes once we have
  compiler support for 64bit integers (soon)
2010-05-25 08:06:14 +00:00
Kees van Reeuwijk
ac14a989b3 Fixed some inconsistent strict typing declarations.
Better strict typing.
2010-05-25 07:23:24 +00:00
David van Moolenbroek
2a8961cdac fix int64.3 man page formatting 2010-05-24 21:10:37 +00:00
Erik van der Kouwe
1f11a57141 Oops, last commit included more than was intended 2010-05-20 08:07:47 +00:00
Erik van der Kouwe
5f15ec05b2 More system processes, this was not enough for the release script to run on some configurations 2010-05-20 08:05:07 +00:00
Erik van der Kouwe
456a125e00 Minor patch: increase size of root partition for minimal release, eve very small additions with -c made it fail 2010-05-19 20:11:22 +00:00
Arun Thomas
b0159ad168 Buildsystem changes for GCC
-Makefile updates
-Update mkdep
-Build fixes/warning cleanups for some programs
-Restore leading underscores on global syms in kernel asm files
-Increase ramdisk size
2010-05-19 13:24:15 +00:00
Ben Gras
bcdaf033b5 pm - fix sched interaction
For coredumping processes, PM forgets to inform SCHED that the
process has vanished, causing future fork()s to fail.
2010-05-19 13:22:29 +00:00
Erik van der Kouwe
77d5243f4a More sensible description for EPERM from http://www.opengroup.org/onlinepubs/000095399/basedefs/errno.h.html 2010-05-19 11:51:49 +00:00