| History log of /src/sys/altq/altq_rmclass.h |
| Revision | | Date | Author | Comments |
| 1.14 |
| 03-Feb-2025 |
ozaki-r | altq, cbq: fix TS_ADD_DELTA on ILP32
TS_ADD_DELTA adds a nanosecond value, delta, to a timespec value while
delta is int64_t and tv_nsec is long. delta is guaranteed as positive
in ALTQ/CBQ. However, the original code used delta after casting it
into long. So it could be negative on ILP32 systems. TS_ADD_DELTA
didn't assume that delta was negative and resulting tv_nsec could be
negative unexpectedly.
This issue was likely to occur when bandwidth borrowing was used where
the borrowing class had a absolutely small bandwidth (e.g., less than
10 Mbps) or a relatively small bandwidth (e.g., 10 Mbps when the parent
class had 200 Mbps).
KASSERT has been added to guarantee resulting tv_nsec is always not
negative. Additionally, another KASSERT has been added to make it clear
that delta is assumed to be non-negative.
|
| 1.13 |
| 24-May-2022 |
andvar | branches: 1.13.4; 1.13.10;
fix various typos in comment, documentation and log messages.
|
| 1.12 |
| 21-Jul-2021 |
ozaki-r | altq, cbq: change long to uint64_t for ILP32 archs
PR kern/56319
|
| 1.11 |
| 21-Jul-2021 |
ozaki-r | altq, cbq: treat time in nanoseconds
Time granularity in microseconds is sometimes not enough to provide
accurate bandwidth control; actual bandwidth on a capped class
can exceed its limit considerably in some cases.
Treating time in nanoseconds requires the following changes:
- Use timespec instead of timeval
- Use nanotime(9) instead of microtime(9)
- Change the type of some variables, especially *idle, to long
|
| 1.10 |
| 21-Jul-2021 |
ozaki-r | altq, cbq: convert ns_per_byte to ps_per_byte
Also the type of variables of it is changed to u_long from int.
This change provides fine-grain resolution of bandwidth. For example
750 Mbps was treated as 800 Mbps internally because bandwidth was
represented as nanoseconds per byte. Converting the representation
to picoseconds per byte enables to treat 750 Mbps as-is.
PR kern/56319
|
| 1.9 |
| 13-Jul-2021 |
ozaki-r | altq, cbq: apply the do-while idiom to TV_* macros
|
| 1.8 |
| 28-Oct-2006 |
peter | branches: 1.8.158;
Don't allow to initialize an interface with MTU smaller than one.
This prevents a divide by zero and fixes PR #21474.
ok tron@
|
| 1.7 |
| 12-Oct-2006 |
peter | Merge the peter-altq branch.
(sync with KAME & add support for using ALTQ with pf(4)).
|
| 1.6 |
| 11-Dec-2005 |
christos | branches: 1.6.12; 1.6.20; 1.6.22;
merge ktrace-lwp.
|
| 1.5 |
| 26-Feb-2005 |
perry | branches: 1.5.4;
nuke trailing whitespace
|
| 1.4 |
| 20-Aug-2003 |
itojun | branches: 1.4.8; 1.4.10;
expose CBQ_MAXPRI to userland
|
| 1.3 |
| 05-Mar-2002 |
itojun | branches: 1.3.14;
bring in latest ALTQ from kjc. ALTQify some of the drivers.
|
| 1.2 |
| 14-Dec-2000 |
thorpej | branches: 1.2.2; 1.2.4; 1.2.6;
Add NetBSD RCS IDs.
|
| 1.1 |
| 14-Dec-2000 |
thorpej | branches: 1.1.1;
Initial revision
|
| 1.1.1.1 |
| 14-Dec-2000 |
thorpej | Import ALTQ from KAME.
|
| 1.2.6.1 |
| 16-Mar-2002 |
jdolecek | Catch up with -current.
|
| 1.2.4.1 |
| 01-Apr-2002 |
nathanw | Catch up to -current.
(CVS: It's not just a program. It's an adventure!)
|
| 1.2.2.2 |
| 05-Jan-2001 |
bouyer | Sync with HEAD
|
| 1.2.2.1 |
| 14-Dec-2000 |
bouyer | file altq_rmclass.h was added on branch thorpej_scsipi on 2001-01-05 17:39:37 +0000
|
| 1.3.14.4 |
| 04-Mar-2005 |
skrll | Sync with HEAD.
Hi Perry!
|
| 1.3.14.3 |
| 21-Sep-2004 |
skrll | Fix the sync with head I botched.
|
| 1.3.14.2 |
| 18-Sep-2004 |
skrll | Sync with HEAD.
|
| 1.3.14.1 |
| 03-Aug-2004 |
skrll | Sync with HEAD
|
| 1.4.10.1 |
| 19-Mar-2005 |
yamt | sync with head. xen and whitespace. xen part is not finished.
|
| 1.4.8.1 |
| 29-Apr-2005 |
kent | sync with -current
|
| 1.5.4.1 |
| 30-Dec-2006 |
yamt | sync with head.
|
| 1.6.22.2 |
| 10-Dec-2006 |
yamt | sync with head.
|
| 1.6.22.1 |
| 22-Oct-2006 |
yamt | sync with head
|
| 1.6.20.1 |
| 18-Nov-2006 |
ad | Sync with head.
|
| 1.6.12.1 |
| 18-Mar-2006 |
peter | Sync with KAME.
|
| 1.8.158.1 |
| 01-Aug-2021 |
thorpej | Sync with HEAD.
|
| 1.13.10.1 |
| 02-Aug-2025 |
perseant | Sync with HEAD
|
| 1.13.4.1 |
| 20-Feb-2025 |
martin | Pull up following revision(s) (requested by ozaki-r in ticket #1049):
sys/altq/altq_rmclass.h: revision 1.14
altq, cbq: fix TS_ADD_DELTA on ILP32
TS_ADD_DELTA adds a nanosecond value, delta, to a timespec value while
delta is int64_t and tv_nsec is long. delta is guaranteed as positive
in ALTQ/CBQ. However, the original code used delta after casting it
into long. So it could be negative on ILP32 systems. TS_ADD_DELTA
didn't assume that delta was negative and resulting tv_nsec could be
negative unexpectedly.
This issue was likely to occur when bandwidth borrowing was used where
the borrowing class had a absolutely small bandwidth (e.g., less than
10 Mbps) or a relatively small bandwidth (e.g., 10 Mbps when the parent
class had 200 Mbps).
KASSERT has been added to guarantee resulting tv_nsec is always not
negative. Additionally, another KASSERT has been added to make it clear
that delta is assumed to be non-negative.
|
