dragonfly.git
15 years agoAdd missing splx(s) in sk_detach.
Joerg Sonnenberger [Thu, 26 May 2005 22:17:12 +0000 (22:17 +0000)]
Add missing splx(s) in sk_detach.

15 years ago* Declare vtb_buffer as a pointer to uint16_t and get rid of
Sascha Wildner [Thu, 26 May 2005 16:24:33 +0000 (16:24 +0000)]
* Declare vtb_buffer as a pointer to uint16_t and get rid of
  sc_vtb_pointer() and some casts.

* Use the standard uint16_t type instead of u_int16_t.

* Use M_ZERO instead of explicitly bzero()ing.

* Don't set VTB_VALID if malloc() failed.

In discussion with: joerg

15 years agoProperly initialize the serializer by calling lwkt_serialize_init().
Matthew Dillon [Thu, 26 May 2005 09:10:36 +0000 (09:10 +0000)]
Properly initialize the serializer by calling lwkt_serialize_init().

15 years agoFix a bug in the serializer's race detection code. It is possible for
Matthew Dillon [Thu, 26 May 2005 09:10:10 +0000 (09:10 +0000)]
Fix a bug in the serializer's race detection code.  It is possible for
the lock to be acquired and released by an interrupt between our attempt
to acquire and our tsleep.  We must test the condition again from within
a critical section before sleeping.

The previous code was doing a test-and-set when it should have just done
a test in the re-test case.

15 years agoRemove some debugging that crept int othe last commit.
Matthew Dillon [Thu, 26 May 2005 09:06:40 +0000 (09:06 +0000)]
Remove some debugging that crept int othe last commit.

15 years agoFix a race in the serializer's use of atomic_intr_cond_enter(). When
Matthew Dillon [Wed, 25 May 2005 22:59:20 +0000 (22:59 +0000)]
Fix a race in the serializer's use of atomic_intr_cond_enter().  When
the procedure call is made the serializer must enter a critical section
and check the bit again before actually sleeping to avoid a race against
wakeup().

15 years agoPrefer m_getcl() to separate calls to MGETHDR() and MCLGET() in order to
Jeffrey Hsu [Wed, 25 May 2005 21:35:51 +0000 (21:35 +0000)]
Prefer m_getcl() to separate calls to MGETHDR() and MCLGET() in order to
take advantage of cluster caching.

15 years agoProperly conditionalize a call to ether_poll_deregister via
Matthew Dillon [Wed, 25 May 2005 21:26:52 +0000 (21:26 +0000)]
Properly conditionalize a call to ether_poll_deregister via
#ifdef DEVICE_POLLING.

Reported-by: various
15 years agoRemove redundant assignment.
Jeffrey Hsu [Wed, 25 May 2005 21:18:51 +0000 (21:18 +0000)]
Remove redundant assignment.

15 years agoGet rid of IFCAP_POLLING for now.
Matthew Dillon [Wed, 25 May 2005 14:59:05 +0000 (14:59 +0000)]
Get rid of IFCAP_POLLING for now.

15 years agoChange ifconfig 'poll' to ifconfig 'polling', the same keyword that it
Matthew Dillon [Wed, 25 May 2005 14:58:32 +0000 (14:58 +0000)]
Change ifconfig 'poll' to ifconfig 'polling', the same keyword that it
used for the capabilities flag before.  But now it just uses a normal
IFF flag.

15 years agoRemove some compat junk, deindent a switch.
Joerg Sonnenberger [Wed, 25 May 2005 13:12:22 +0000 (13:12 +0000)]
Remove some compat junk, deindent a switch.

15 years agoConvert to new m_ext API.
Joerg Sonnenberger [Wed, 25 May 2005 12:37:29 +0000 (12:37 +0000)]
Convert to new m_ext API.

15 years agoForce jumbo buffers to be a multiple of 64bit.
Joerg Sonnenberger [Wed, 25 May 2005 12:18:31 +0000 (12:18 +0000)]
Force jumbo buffers to be a multiple of 64bit.

15 years agoRemove compat junk.
Joerg Sonnenberger [Wed, 25 May 2005 11:59:14 +0000 (11:59 +0000)]
Remove compat junk.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoUse ether_crc32_le.
Joerg Sonnenberger [Wed, 25 May 2005 11:51:26 +0000 (11:51 +0000)]
Use ether_crc32_le.

Obtained-from: FreeBSD

Don't include machine/clock.h, we have DELAY already.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoif_printf / device_printf and some more cleanup.
Joerg Sonnenberger [Wed, 25 May 2005 11:46:10 +0000 (11:46 +0000)]
if_printf / device_printf and some more cleanup.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoNuke further compatibility junk.
Joerg Sonnenberger [Wed, 25 May 2005 11:42:59 +0000 (11:42 +0000)]
Nuke further compatibility junk.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoif_printf / device_printf and some further cleanup.
Joerg Sonnenberger [Wed, 25 May 2005 11:38:59 +0000 (11:38 +0000)]
if_printf / device_printf and some further cleanup.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoFix a bugaboo in the last commit. We can't move to a generic mux yet
Matthew Dillon [Wed, 25 May 2005 07:58:41 +0000 (07:58 +0000)]
Fix a bugaboo in the last commit.  We can't move to a generic mux yet
because FAST interrupts aren't compatible with the mechanism.

15 years agoDon't treat an empty command buffer as parse error.
YONETANI Tomokazu [Wed, 25 May 2005 06:22:22 +0000 (06:22 +0000)]
Don't treat an empty command buffer as parse error.
The buffer will also be made empty after /boot.config
is successfully parsed, or when you hit just Enter key
to suppress 5-second wait.

15 years agoRewrite the polling code. Instead of trying to do fancy polling enablement
Matthew Dillon [Wed, 25 May 2005 01:44:33 +0000 (01:44 +0000)]
Rewrite the polling code.  Instead of trying to do fancy polling enablement
from inside the IF interrupt itself, which creates a headache in the code,
simply allow IFF_POLLING to be set and cleared via ifconfig.  This greatly
simplifies both the networking code and the polling code and allows polling
to be enabled and disabled at will on a per-network-interface basis.

* Drivers no longer have to have polling checks in the interrupt path.
* An if_poll function vector has been added.  Polling is supported if the
  driver initializes the vector.
* Registration command added to the poll function command list.
* Driver code for registration and deregistration is now greatly simplified.

The kernel polling code no longer randomly turns off the polling bit if an
interface goes down or is reset.

Remove IFCAP_POLLING, it serves no purpose.

Fix a couple of bugs in the serializer code.  Add a warning in
nexus_setup_intr if a driver tries to specify a serializer and an SPL.
A driver can specify one or the other, not both.

Convert the EM driver to use the new serializer API instead of SPLs.

Add ifconfig poll and ifconfig -poll support to ifconfig, and fix bugs
in the rtsock code that only returned the low 16 bits of the interface
flags so ifconfig properly reports when polling mode is turned on for an
interface.

NOTE to people using polling.  You must first enable polling via
kern.polling.enable, and then may specify the 'poll' directive in ifconfig
to enable it on a per interface basis.  If IFF_POLLING refuses to be set,
the device does not support polling.

15 years ago(add missing file related to last commit)
Matthew Dillon [Tue, 24 May 2005 21:22:05 +0000 (21:22 +0000)]
(add missing file related to last commit)

Get rid of bus_{disable,enable}_intr(), it wasn't generic enough for
our needs.

Implement some generic atomic.h functions to aid in the implementation of
a low level mutex.

Implement a generic low level sleep-mutex serializer, kern/lwkt_serialize.c.
The serializer is designed to be a replacement for SPL calls but may also
be used for other very low level work (e.g. lockmgr interlocks).

Add a serializer argument to BUS_SETUP_INTR().  When non-NULL, the interrupt
handler will no longer be protected by an SPL so e.g. spl*() will no
longer protect against that device's interrupts.

The IF queueing and dequeueing mechanisms may no longer depend on outside
SPL state because network driver interrupt handlers are no longer required to
enter splnet().  Use critical sections for the moment.  The IFQ and
IFF_OACTIVE interactions are not yet MP safe.

15 years agoGet rid of the bad hack that was doing network polling from the trap code.
Matthew Dillon [Tue, 24 May 2005 21:18:27 +0000 (21:18 +0000)]
Get rid of the bad hack that was doing network polling from the trap code.

15 years agoGet rid of bus_{disable,enable}_intr(), it wasn't generic enough for
Matthew Dillon [Tue, 24 May 2005 20:59:05 +0000 (20:59 +0000)]
Get rid of bus_{disable,enable}_intr(), it wasn't generic enough for
our needs.

Implement some generic atomic.h functions to aid in the implementation of
a low level mutex.

Implement a generic low level sleep-mutex serializer, kern/lwkt_serialize.c.
The serializer is designed to be a replacement for SPL calls but may also
be used for other very low level work (e.g. lockmgr interlocks).

Add a serializer argument to BUS_SETUP_INTR().  When non-NULL, the interrupt
handler will no longer be protected by an SPL so e.g. spl*() will no
longer protect against that device's interrupts.

The IF queueing and dequeueing mechanisms may no longer depend on outside
SPL state because network driver interrupt handlers are no longer required to
enter splnet().  Use critical sections for the moment.  The IFQ and
IFF_OACTIVE interactions are not yet MP safe.

15 years agoUse ether_crc32_be.
Joerg Sonnenberger [Tue, 24 May 2005 16:46:26 +0000 (16:46 +0000)]
Use ether_crc32_be.

Obtained-from: FreeBSD

15 years agostyle
Joerg Sonnenberger [Tue, 24 May 2005 16:44:41 +0000 (16:44 +0000)]
style

15 years agoConvert to new m_ext API.
Joerg Sonnenberger [Tue, 24 May 2005 11:42:07 +0000 (11:42 +0000)]
Convert to new m_ext API.

15 years agoUse bus_alloc_resource_any when possible.
Joerg Sonnenberger [Tue, 24 May 2005 09:52:15 +0000 (09:52 +0000)]
Use bus_alloc_resource_any when possible.

15 years agoFreeBSD-Date: 2005/05/23 14:27:52
Max Okumoto [Tue, 24 May 2005 09:27:34 +0000 (09:27 +0000)]
FreeBSD-Date: 2005/05/23 14:27:52
Author: harti

Make the ArgArray_Init() function visible to the rest of make.
It will soon be needed for the .SHELL parsing code.

15 years agoif_printf / device_printf.
Joerg Sonnenberger [Tue, 24 May 2005 08:16:07 +0000 (08:16 +0000)]
if_printf / device_printf.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoAdd a device ID for MPI350.
Joerg Sonnenberger [Tue, 24 May 2005 08:10:33 +0000 (08:10 +0000)]
Add a device ID for MPI350.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoUse ether_crc32_le.
Joerg Sonnenberger [Tue, 24 May 2005 07:56:30 +0000 (07:56 +0000)]
Use ether_crc32_le.

Obtained-from: FreeBSD

15 years agoNo machine/clock.h needed.
Joerg Sonnenberger [Tue, 24 May 2005 07:54:49 +0000 (07:54 +0000)]
No machine/clock.h needed.

15 years agoNo FreeBSD 5/6 support here.
Joerg Sonnenberger [Tue, 24 May 2005 07:54:18 +0000 (07:54 +0000)]
No FreeBSD 5/6 support here.

15 years agoif_printf / device_printf.
Joerg Sonnenberger [Tue, 24 May 2005 07:36:29 +0000 (07:36 +0000)]
if_printf / device_printf.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoUse ether_crc32_be.
Joerg Sonnenberger [Tue, 24 May 2005 07:26:41 +0000 (07:26 +0000)]
Use ether_crc32_be.

Obtained-from: FreeBSD

15 years agoRevert to old local unit allocation until I find a better solution
Joerg Sonnenberger [Tue, 24 May 2005 07:25:06 +0000 (07:25 +0000)]
Revert to old local unit allocation until I find a better solution
to unbreak multiport cards.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoMerge from vendor branch GCC:
Joerg Sonnenberger [Tue, 24 May 2005 06:35:19 +0000 (06:35 +0000)]
Merge from vendor branch GCC:
Update for GCC 3.4.4.  Note that the prefered way to fetch just
the core, g++, g77 and objc part. ada, java and the testsuite
are not used, so you don't have to fetch it ever.

15 years agoUpdate for GCC 3.4.4. Note that the prefered way to fetch just
Joerg Sonnenberger [Tue, 24 May 2005 06:35:19 +0000 (06:35 +0000)]
Update for GCC 3.4.4.  Note that the prefered way to fetch just
the core, g++, g77 and objc part. ada, java and the testsuite
are not used, so you don't have to fetch it ever.

15 years agoUpdate in-tree GCC to 3.4.4.
Joerg Sonnenberger [Tue, 24 May 2005 06:29:33 +0000 (06:29 +0000)]
Update in-tree GCC to 3.4.4.

15 years agoUpdate in-tree GCC to 3.4.4.
Joerg Sonnenberger [Tue, 24 May 2005 06:29:33 +0000 (06:29 +0000)]
Update in-tree GCC to 3.4.4.

15 years agoMerge from vendor branch GCC:
Joerg Sonnenberger [Tue, 24 May 2005 06:29:33 +0000 (06:29 +0000)]
Merge from vendor branch GCC:
Update in-tree GCC to 3.4.4.

15 years agoAdd support for dntpd to /etc/defaults/rc.d
Robert Garrett [Tue, 24 May 2005 02:53:13 +0000 (02:53 +0000)]
Add support for dntpd to /etc/defaults/rc.d

15 years agoAdd support for dntpd to /etc/rc.d
Robert Garrett [Tue, 24 May 2005 02:51:49 +0000 (02:51 +0000)]
Add support for dntpd to /etc/rc.d

submitted by rumko

15 years agoMove parseIncPath and sysIncPath out of global scope
Max Okumoto [Mon, 23 May 2005 20:05:05 +0000 (20:05 +0000)]
Move parseIncPath and sysIncPath out of global scope

15 years agoMove the create list variable out of global scope and put it into
Max Okumoto [Mon, 23 May 2005 20:04:43 +0000 (20:04 +0000)]
Move the create list variable out of global scope and put it into
the newly created Parser object.  And Pass the parser object around
as a function paramter.

15 years agoo Remove out of date comment about Compat_Init()
Max Okumoto [Mon, 23 May 2005 20:04:04 +0000 (20:04 +0000)]
o Remove out of date comment about Compat_Init()
o Move init code into Make_Run()
o Have Compat_Run() always return a boolean value (FALSE) so it has
  same prototype as Make_Run().

15 years agoRemove spurious semicolon.
Joerg Sonnenberger [Mon, 23 May 2005 19:45:19 +0000 (19:45 +0000)]
Remove spurious semicolon.

15 years agoFake DES was removed a while ago, remove the prototypes as well.
Joerg Sonnenberger [Mon, 23 May 2005 19:41:59 +0000 (19:41 +0000)]
Fake DES was removed a while ago, remove the prototypes as well.

15 years agoPass queryFlag as a function paramter instead of a program
Max Okumoto [Mon, 23 May 2005 18:26:25 +0000 (18:26 +0000)]
Pass queryFlag as a function paramter instead of a program
global variable.

15 years agoPass ENDNode as a function parameter instead of using a file global variable.
Max Okumoto [Mon, 23 May 2005 18:25:58 +0000 (18:25 +0000)]
Pass ENDNode as a function parameter instead of using a file global variable.

15 years agoMove signal handler setup code into its own function, and share
Max Okumoto [Mon, 23 May 2005 18:25:34 +0000 (18:25 +0000)]
Move signal handler setup code into its own function, and share
that with the compat code.

15 years agoMove ProcExec() into proc.c
Max Okumoto [Mon, 23 May 2005 18:24:59 +0000 (18:24 +0000)]
Move ProcExec() into proc.c

15 years agoDo better range checking on the LDT. FreeBSD-SA-05:07.ldt
Matthew Dillon [Mon, 23 May 2005 18:23:29 +0000 (18:23 +0000)]
Do better range checking on the LDT.  FreeBSD-SA-05:07.ldt

15 years agoduplicate is
Joerg Sonnenberger [Mon, 23 May 2005 18:22:41 +0000 (18:22 +0000)]
duplicate is

15 years agoHave the EM device call the new interrupt enablement and disablement
Matthew Dillon [Mon, 23 May 2005 18:20:47 +0000 (18:20 +0000)]
Have the EM device call the new interrupt enablement and disablement
functions to avoid handler races.

15 years agoAdd NEWBUS infrastructure for interrupt enablement and disablement. This
Matthew Dillon [Mon, 23 May 2005 18:19:55 +0000 (18:19 +0000)]
Add NEWBUS infrastructure for interrupt enablement and disablement.  This
allows a device to indicate to the interrupt dispatch architecture that it
has enabled or disabled the device interrupt at the source.  The dispatch
will then decline to call the handler.   This is necessary because it is
possible for the interrupt handler to be called from the interrupt thread
AFTER the device has disabled the hard interrupt.   There are two cases:

    FIRST CASE:

* hard interrupt occurs
* interrupt thread is scheduled but cannot preempt
* device disables interrupt
* interrupt thread then runs handler while device believes interrupt to
  be disabled.

    SECOND CASE:

* multiple devices share the same interrupt (#1 and #2)
* device #1 interrupts and schedules the thread
* the handler for ALL devices is run, even if device #2 disabled
  its hard interrupt.

Clean up and simplify the interrupt vector code.  Always install a MUX
function.  The MUX function will check the handler enablement/disablement
state.

15 years agoSome comment header and whitespace cleanups.
Max Okumoto [Mon, 23 May 2005 18:19:05 +0000 (18:19 +0000)]
Some comment header and whitespace cleanups.

15 years agodevice_printf / if_printf cleanup.
Joerg Sonnenberger [Mon, 23 May 2005 18:05:58 +0000 (18:05 +0000)]
device_printf / if_printf cleanup.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoCleanup bfe_probe.
Joerg Sonnenberger [Mon, 23 May 2005 17:54:11 +0000 (17:54 +0000)]
Cleanup bfe_probe.

15 years agostyle(9) and nic style changes.
Joerg Sonnenberger [Mon, 23 May 2005 16:00:44 +0000 (16:00 +0000)]
style(9) and nic style changes.

15 years agoSpurious semicolon broke gcc2 build.
Joerg Sonnenberger [Mon, 23 May 2005 07:00:36 +0000 (07:00 +0000)]
Spurious semicolon broke gcc2 build.

15 years ago- util.h -> libutil.h
Liam J. Foy [Sun, 22 May 2005 17:48:57 +0000 (17:48 +0000)]
- util.h -> libutil.h

Requested by: Joerg@

15 years ago- More fixes for the debug code
Liam J. Foy [Sun, 22 May 2005 17:09:53 +0000 (17:09 +0000)]
- More fixes for the debug code

15 years agoConvert bge(4) to the new M_EXT API. This allows merging the dynamic
Joerg Sonnenberger [Sun, 22 May 2005 16:14:04 +0000 (16:14 +0000)]
Convert bge(4) to the new M_EXT API. This allows merging the dynamic
state back into the softc and saves a small malloc per jumbo buffer
in the rx ring.

15 years ago- WITHOUT_ERRNO should be 0, silly mistake.
Liam J. Foy [Sun, 22 May 2005 11:53:05 +0000 (11:53 +0000)]
- WITHOUT_ERRNO should be 0, silly mistake.

Noticed by: swildner@

15 years agoMore -Wwrite-strings cleanup and make sure you can actually play it.
YONETANI Tomokazu [Sun, 22 May 2005 03:37:05 +0000 (03:37 +0000)]
More -Wwrite-strings cleanup and make sure you can actually play it.

15 years agoMinimum bandaid to the breakage due to the last gcc40-cleanup.
YONETANI Tomokazu [Sun, 22 May 2005 01:28:15 +0000 (01:28 +0000)]
Minimum bandaid to the breakage due to the last gcc40-cleanup.

15 years agoAdd -t/--tag option to automatically add $DragonFly$ tags to the patch files
Simon Schubert [Sat, 21 May 2005 10:47:43 +0000 (10:47 +0000)]
Add -t/--tag option to automatically add $DragonFly$ tags to the patch files

15 years agoFix compilation with gcc40: Don't postincrement and cast lvalues.
Simon Schubert [Sat, 21 May 2005 10:31:08 +0000 (10:31 +0000)]
Fix compilation with gcc40: Don't postincrement and cast lvalues.

15 years agoUse ether_crc32_le instead of local hack.
Joerg Sonnenberger [Sat, 21 May 2005 09:07:52 +0000 (09:07 +0000)]
Use ether_crc32_le instead of local hack.

15 years agostyle(9)
Joerg Sonnenberger [Sat, 21 May 2005 09:05:05 +0000 (09:05 +0000)]
style(9)

15 years agoExploit bus_alloc_resource_any.
Joerg Sonnenberger [Sat, 21 May 2005 08:57:29 +0000 (08:57 +0000)]
Exploit bus_alloc_resource_any.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoUse bus_alloc_resource_any. Use pci helper functions, don't roll them
Joerg Sonnenberger [Sat, 21 May 2005 07:38:41 +0000 (07:38 +0000)]
Use bus_alloc_resource_any. Use pci helper functions, don't roll them
by hand.

15 years agoif_printf/device_printf cleanup. Remove minor junk.
Joerg Sonnenberger [Sat, 21 May 2005 07:28:04 +0000 (07:28 +0000)]
if_printf/device_printf cleanup. Remove minor junk.

Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
15 years agoRework TX EOF handling. We have to always check for TX underruns,
Joerg Sonnenberger [Fri, 20 May 2005 14:30:33 +0000 (14:30 +0000)]
Rework TX EOF handling. We have to always check for TX underruns,
even in the TX_OK case. In the error case, reset the tx configuration
register.

Inspired-by: NetBSD'S rtl81x9 driver
15 years agoCleanup str_concat() there is no need for special flags. Just
Max Okumoto [Fri, 20 May 2005 11:48:55 +0000 (11:48 +0000)]
Cleanup str_concat() there is no need for special flags.  Just
pass the character you want to use as a seperator.  If the seperator
character is '\0', then don't put in a seperator.

15 years agoGet rid of shellPath global by adding another field to the
Max Okumoto [Fri, 20 May 2005 11:48:18 +0000 (11:48 +0000)]
Get rid of shellPath global by adding another field to the
shell structure.

15 years agoRename functions since they are no longer part of job.c
Max Okumoto [Fri, 20 May 2005 11:47:43 +0000 (11:47 +0000)]
Rename functions since they are no longer part of job.c

15 years agoRemove shells array and just use the const strings directly. This
Max Okumoto [Fri, 20 May 2005 11:47:22 +0000 (11:47 +0000)]
Remove shells array and just use the const strings directly.  This
allows us to get rid of the DEF_SHELL_STRUCT() macro, and just use
a normal struct.

15 years agoRemove string array called 'copy' and just use the const strings directly.
Max Okumoto [Fri, 20 May 2005 11:46:27 +0000 (11:46 +0000)]
Remove string array called 'copy' and just use the const strings directly.

15 years agoAdd comments to shells array definition.
Max Okumoto [Thu, 19 May 2005 17:11:08 +0000 (17:11 +0000)]
Add comments to shells array definition.

15 years agoFreeBSD-Date: 2005/05/18 07:50:38
Max Okumoto [Thu, 19 May 2005 17:10:37 +0000 (17:10 +0000)]
FreeBSD-Date: 2005/05/18 07:50:38
Author: harti

Get rid of the ReturnStatus obscuration that was anyway used only
in two places. While here don't bother returning anything from
Lst_Replace - nobody ever checks the return code.

Suggested by:   jmallet

15 years agoAdd the missing files. <sigh>
Max Okumoto [Thu, 19 May 2005 17:09:20 +0000 (17:09 +0000)]
Add the missing files.  <sigh>

15 years ago2nd Try. The correct version.
Max Okumoto [Thu, 19 May 2005 17:07:17 +0000 (17:07 +0000)]
2nd Try.  The correct version.

Move shell code out of job.c into shell.[ch]

15 years agoRevert patch on 2005/05/19 17:53:58 it was the wrong one :-(
Max Okumoto [Thu, 19 May 2005 17:04:45 +0000 (17:04 +0000)]
Revert patch on 2005/05/19 17:53:58 it was the wrong one :-(

15 years agoBackout last commit. It should be done after the next patch. <sigh>
Max Okumoto [Thu, 19 May 2005 16:56:37 +0000 (16:56 +0000)]
Backout last commit.  It should be done after the next patch. <sigh>

15 years agoAdd comments to shells array definition.
Max Okumoto [Thu, 19 May 2005 16:54:15 +0000 (16:54 +0000)]
Add comments to shells array definition.

15 years agoMove shell code out of job.c into shell.[ch]
Max Okumoto [Thu, 19 May 2005 16:53:58 +0000 (16:53 +0000)]
Move shell code out of job.c into shell.[ch]

15 years agoMove RLIMIT_NOFILE code into Proc_Init()
Max Okumoto [Thu, 19 May 2005 16:51:45 +0000 (16:51 +0000)]
Move RLIMIT_NOFILE code into Proc_Init()
Move check_make_level() into job.c

15 years agoSmall cleanups
Max Okumoto [Thu, 19 May 2005 16:51:19 +0000 (16:51 +0000)]
Small cleanups

15 years agoMove make global variables setup into a sepearate function.
Max Okumoto [Thu, 19 May 2005 16:51:06 +0000 (16:51 +0000)]
Move make global variables setup into a sepearate function.

15 years agoGroup code that sets global variables (make MACROs)
Max Okumoto [Thu, 19 May 2005 16:50:35 +0000 (16:50 +0000)]
Group code that sets global variables (make MACROs)

15 years agoMove code that reads in the input files into sepearte function.
Max Okumoto [Thu, 19 May 2005 16:49:56 +0000 (16:49 +0000)]
Move code that reads in the input files into sepearte function.

15 years agoMove a few global variables into a struct and pass it as a parameter.
Max Okumoto [Thu, 19 May 2005 16:49:32 +0000 (16:49 +0000)]
Move a few global variables into a struct and pass it as a parameter.

15 years agoCleanup search for default Makefile, by adding comments and
Max Okumoto [Thu, 19 May 2005 16:47:57 +0000 (16:47 +0000)]
Cleanup search for default Makefile, by adding comments and
retructuring code.

15 years agoSpelling error caught by harti.
Max Okumoto [Wed, 18 May 2005 21:35:17 +0000 (21:35 +0000)]
Spelling error caught by harti.

15 years agoReduce diff from FreeBSD.
Max Okumoto [Wed, 18 May 2005 07:11:01 +0000 (07:11 +0000)]
Reduce diff from FreeBSD.

15 years ago- Actually make the debug code work
Liam J. Foy [Tue, 17 May 2005 18:09:44 +0000 (18:09 +0000)]
- Actually make the debug code work
- WARNS 6 -DEBUG

Ok boy: Joerg@