| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2003-05-20 | fix a breakage in hfsc. | Kenjiro Cho | |
| hif->hif_rootclass should be initialized when the root queue is created. | |||
| 2003-05-12 | ansi | Henning Brauer | |
| 2003-04-12 | life's easier if one initializes pointers... | Henning Brauer | |
| 2003-04-12 | don't create a root class implicitely. we do this from pfctl, and need | Henning Brauer | |
| control over the root class from pfctl. also fix a few null pointer derefs in case a root class does not exist | |||
| 2003-04-12 | kill the qid assignment code here; return EINVAL if qid == 0 - similar to | Henning Brauer | |
| what we do for teh other schedulers already | |||
| 2003-04-12 | ansi and a bit style; ok kjc@ | Henning Brauer | |
| 2003-04-03 | until now, the queue ID and the priority were tied together with PRIQ. this | Henning Brauer | |
| diff changes that. with PRIQ, the queues are in an array, with the priority as key. removing the tie means we cannot access the array with (queueID - 1) as key any more but need to traverse the array until the queue ID matches. As the array has a maximum of 16 entries, traversing linear is okay. a new queue ID allocation algorithm coming soon will require this. ok dhartmei@ kjc@ | |||
| 2003-04-03 | remove remaining lines not needed by openbsd. | Kenjiro Cho | |
| 2003-04-02 | remove dead code | Henning Brauer | |
| 2003-04-02 | remove what is left over from the control class stuff | Henning Brauer | |
| ok kjc@ | |||
| 2003-03-31 | remove queue ID assigning code. it's dead since some time as we assign queue | Henning Brauer | |
| IDs in userland already. | |||
| 2003-03-27 | double ; | Henning Brauer | |
| Patrick Latifi, Thanks! | |||
| 2003-03-27 | idx is unsigned | Henning Brauer | |
| from Patrick Latifi | |||
| 2003-03-24 | fix error return values (NULL to ENOMEM) and add missing ellist_destroy() | Kenjiro Cho | |
| in error handling of hfsc_add_altq(). report and fix by millert@ | |||
| 2003-03-13 | scale the red thresholds according to the queue limit. | Kenjiro Cho | |
| the min and max thresholds are set to 10% and 30% of the queue limit. | |||
| 2003-03-13 | remove an unused function prototype. | Kenjiro Cho | |
| from KAME | |||
| 2003-03-11 | add protection against packets without pkthdr. | Kenjiro Cho | |
| this should not happen but just in case. printf() is intended to be annoying so that we'll get reports on it. original idea from dhartmei@ ok deraadt@, henning@ | |||
| 2003-03-02 | only assign a new queue id if a->qid is 0. otherwise, check a->qid for | Henning Brauer | |
| validity and take that. validity checks provided by kjc@ ok dhartmei@ kjc@ | |||
| 2003-02-08 | ansi | Henning Brauer | |
| 2003-02-05 | missed one during ansification | Henning Brauer | |
| 2003-01-30 | big KNF & ANSI round | Henning Brauer | |
| "accidently" typed while hunting a bug ;-) ok kjc@ | |||
| 2003-01-07 | remove the altq classifier code which is replaced by pf and no longer used. | Kenjiro Cho | |
| ok henning@, deraadt@ | |||
| 2003-01-07 | Correct several typos in comments. No code changes. | Chad Loder | |
| 2002-12-18 | enable PRIQ and HFSC if ALTQ itself is enabled. | Henning Brauer | |
| ok theo | |||
| 2002-12-16 | major KNF, Take 2 | Henning Brauer | |
| ok kjc@ | |||
| 2002-12-16 | switchover to pf-based altq. | Kenjiro Cho | |
| - remove files which are no longer used, or we don't have plans to support in pf in the near future. - remove altq ioctl related stuff. - convert the PRIQ, HFSC and RIO modules to pf-based altq. (these are not enabled in GENERIC, CDNR is not converted yet.) | |||
| 2002-11-29 | supress printf's. | Kenjiro Cho | |
| 2002-11-29 | convert read_machclk() from a macro to a real function and | Kenjiro Cho | |
| avoid kernel-rebuild with ALTQ_NOPCC for non-pentium users. | |||
| 2002-11-26 | fix "pfctl -Fq". | Kenjiro Cho | |
| after altq gets flushed, altq forgot that it was enabled since altq is actually detached with an empty ruleset. so, add a variable, pfaltq_running, to remember the running state and re-enable altq when a new ruleset is loaded. noticed, tested, and oked by henning@ | |||
| 2002-11-26 | KNF | Henning Brauer | |
| 2002-10-11 | make CBQ aware of the pf API so that pf developers can play with CBQ. | Kenjiro Cho | |
| for now, implemnet a tentative transition mechanism to allow the old API and the new API to coexit. the old API will go away when pfctl(8) becomes ready. | |||
| 2002-10-10 | add a borrow flag to be used by pf/altq. | Kenjiro Cho | |
| 2002-10-08 | the first step of pf/altq merge. | Kenjiro Cho | |
| this commit is to allow further development in both userland and kernel. the goal is to replace altq's classifier by pf(4). - make pf tag a queue id to mbuf and make altq read the queue id - merge altq config into pf.conf(5) ok dhartmei@, henning@ | |||
| 2002-10-05 | fix a embarrassing list-handling bug in cbq. | Kenjiro Cho | |
| this bug is triggered only when 3 or more cbq instances are created at a time. | |||
| 2002-07-25 | support 486 which lack the pentium_mhz stuff; from rees, kjc ok | Theo de Raadt | |
| 2002-06-14 | spelling; from Brian Poole <raj@cerias.purdue.edu> | Todd T. Fries | |
| 2002-05-31 | define ALTQ_NOPCC if i386 and !I586_CPU and !I686_CPU. | Kenjiro Cho | |
| otherwise, kernel fails to compile. pointed out and ok by fgsch@ | |||
| 2002-05-17 | sync with KAME. | Kenjiro Cho | |
| - don't MALLOC() with M_WAITOK in the spl block. move the allocation before splimp(). - when we reset vt of a class, reset the runtime service curve as well. - don't use max() to compare 64 bit values. | |||
| 2002-05-17 | sync with KAME. | Kenjiro Cho | |
| update ECN in ALTQ from RFC2481 to RFC3168. | |||
| 2002-03-15 | Cosmetic changes only, primarily making comments line up nicely after the | Todd C. Miller | |
| __P removal. | |||
| 2002-03-14 | Final __P removal plus some cosmetic fixups | Todd C. Miller | |
| 2002-03-14 | First round of __P removal in sys | Todd C. Miller | |
| 2002-03-01 | sync with KAME | Kenjiro Cho | |
| move the function prototype of altq_etherclassify() from netinet/if_ether.h to altq/if_altq.h. | |||
| 2002-02-13 | sync with KAME. | Kenjiro Cho | |
| check if the CPU supports TSC in i386 to avoid panic on non-pentium CPUs. use pentium_mhz or alpha's cycles_per_usec. add a protection against a bogus IP header length when parsing the header. be more careful about mbuf data length. a backtrace reported by "Robert John Shepherd" <robert@reviewer.co.uk> | |||
| 2002-02-13 | sync with KAME. | Kenjiro Cho | |
| comment out the warning message for the clock going backwards in the TV_DELTA macro. it happens when the clock is adjusted and gives me more trouble than help. | |||
| 2002-02-13 | move ALTQ_RED and ALTQ_CBQ from altq.h to altq_var.h. | Kenjiro Cho | |
| protect them from duplicated defines. | |||
| 2001-11-06 | Replace inclusion of <vm/foo.h> with the correct <uvm/bar.h> when necessary. | Miod Vallat | |
| (Look ma, I might have broken the tree) | |||
| 2001-10-26 | avoid divide-by-zero when the specified bandwidth is less than 8bps. | Kenjiro Cho | |
| PR kernel/2150 | |||
| 2001-09-07 | fix IPv4 fragment handling in the classifier. | Kenjiro Cho | |
| report and fix by "Paul Koster" <list@kosteronline.net> | |||
| 2001-08-16 | state that this skeleton file is in the public domain. | Kenjiro Cho | |
 |
index : src | |
| OpenBSD base system |
| summaryrefslogtreecommitdiff |