From 5203edcdc553fda6caa1da8826a89b1a02dad1bf Mon Sep 17 00:00:00 2001 From: Ruslan Ermilov Date: Sat, 3 Jul 2004 18:29:24 +0000 Subject: Mechanically kill hard sentence breaks and double whitespaces. --- share/man/man4/man4.i386/acpi_asus.4 | 3 +- share/man/man4/man4.i386/amdpm.4 | 4 +- share/man/man4/man4.i386/apm.4 | 33 ++++++++----- share/man/man4/man4.i386/ar.4 | 8 ++-- share/man/man4/man4.i386/cs.4 | 24 ++++++---- share/man/man4/man4.i386/ct.4 | 9 ++-- share/man/man4/man4.i386/ex.4 | 6 ++- share/man/man4/man4.i386/fe.4 | 6 ++- share/man/man4/man4.i386/io.4 | 3 +- share/man/man4/man4.i386/mse.4 | 6 ++- share/man/man4/man4.i386/ndis.4 | 36 ++++++++++----- share/man/man4/man4.i386/perfmon.4 | 27 +++++++---- share/man/man4/man4.i386/ray.4 | 8 ++-- share/man/man4/man4.i386/scd.4 | 3 +- share/man/man4/man4.i386/spkr.4 | 39 ++++++++++------ share/man/man4/man4.i386/streams.4 | 2 +- share/man/man4/man4.i386/vx.4 | 6 ++- share/man/man4/man4.i386/wl.4 | 90 ++++++++++++++++++++++++------------ 18 files changed, 207 insertions(+), 106 deletions(-) (limited to 'share/man/man4/man4.i386') diff --git a/share/man/man4/man4.i386/acpi_asus.4 b/share/man/man4/man4.i386/acpi_asus.4 index 2d5314d92165..0e2cbfd469cd 100644 --- a/share/man/man4/man4.i386/acpi_asus.4 +++ b/share/man/man4/man4.i386/acpi_asus.4 @@ -81,7 +81,8 @@ TV-Out .Pp Some models also support video switching via the generic .Xr acpi_video 4 -driver. Most models don't, however. +driver. +Most models don't, however. .El .Pp Defaults for these variables can be set in diff --git a/share/man/man4/man4.i386/amdpm.4 b/share/man/man4/man4.i386/amdpm.4 index 400f79be848f..df513f056453 100644 --- a/share/man/man4/man4.i386/amdpm.4 +++ b/share/man/man4/man4.i386/amdpm.4 @@ -55,7 +55,9 @@ The driver first appeared in .Fx 4.5 . .Sh AUTHORS -This driver was written by Matthew C. Forman. +.An -nosplit +This driver was written by +.An "Matthew C. Forman" . Based heavily on the .Nm alpm driver by diff --git a/share/man/man4/man4.i386/apm.4 b/share/man/man4/man4.i386/apm.4 index ec44015196cf..e9c78b343fca 100644 --- a/share/man/man4/man4.i386/apm.4 +++ b/share/man/man4/man4.i386/apm.4 @@ -40,14 +40,17 @@ comprising of system wakeup time and elapsed time during suspended mode. .It .Nm slows CPU clock when there are no system activities (runnable processes, -interrupts, etc.). This function is available only on systems whose APM +interrupts, etc.). +This function is available only on systems whose APM supports CPU idling. .It .Nm -exports an application interface as a character device. Applications +exports an application interface as a character device. +Applications can control APM, or retrieve APM status information via this interface. .Nm -exports the following interfaces. These symbols are defined in +exports the following interfaces. +These symbols are defined in .In machine/apm_bios.h . .Bl -tag -width 4n -offset indent .It Sy APMIO_SUSPEND @@ -67,11 +70,13 @@ Some APM implementations execute the HLT (Halt CPU until an interrupt occurs) instruction in the .Dq Em Idle CPU -call, while others do not. Thus enabling this may result in +call, while others do not. +Thus enabling this may result in redundant HLT executions because .Dq Em Idle CPU is called from the kernel context switch routine that inherently executes -HLT. This may reduce peak system performance. +HLT. +This may reduce peak system performance. .Pp Also the system hangs up if HLT instruction is disabled in the kernel context switch routine, and if the APM implementation of the machine @@ -89,7 +94,8 @@ The current version of does not call .Dq Em Idle CPU from the kernel context switch routine if clock slowdown is not supported, -and it executes HLT instruction by default. Therefore, there is +and it executes HLT instruction by default. +Therefore, there is no need to use these two operations in most cases. .El .Pp @@ -112,10 +118,14 @@ polls APM events and handles the following events. .El .El .Sh BUGS -WARNING! Many, if not most, of the implementations of APM-bios in laptops -today are buggy. You may be putting your LCD-display and batteries at -a risk by using this interface. (The reason this isn't a problem for -MS-windows is that they use the real-mode interface.) If you see any +WARNING! +Many, if not most, of the implementations of APM-bios in laptops +today are buggy. +You may be putting your LCD-display and batteries at +a risk by using this interface. +(The reason this isn't a problem for +MS-windows is that they use the real-mode interface.) +If you see any weird behavior from your system with this code in use, unplug the power and batteries ASAP, if not immediately, and disable this code. .Pp @@ -129,7 +139,8 @@ may cause serious trouble when resuming the system. BIOS setup programs should be called during bootstrap, or from DOS. .Pp Some APM implementations cannot handle events such as pushing the -power button or closing the cover. On such implementations, the system +power button or closing the cover. +On such implementations, the system .Ar must be suspended .Ar only diff --git a/share/man/man4/man4.i386/ar.4 b/share/man/man4/man4.i386/ar.4 index f2e9c38d4611..114bbc752240 100644 --- a/share/man/man4/man4.i386/ar.4 +++ b/share/man/man4/man4.i386/ar.4 @@ -61,9 +61,11 @@ Alternately, the driver can be compiled to support (see below). .Sh NUMBERING Only one line for each card is needed in the kernel configuration file. -The first card's ports will be installed from ar0. The numbering of the -next card will continue where the first stopped, eg. if the first card -is a two port card it will use ar0 and ar1. The next card will then +The first card's ports will be installed from ar0. +The numbering of the +next card will continue where the first stopped, e.g.\& if the first card +is a two port card it will use ar0 and ar1. +The next card will then start at ar2. .Pp The card only supports IRQ 3, 5, 7, 10, 11, 12 and 15. diff --git a/share/man/man4/man4.i386/cs.4 b/share/man/man4/man4.i386/cs.4 index 9e1efc57dd74..1687a3234ca2 100644 --- a/share/man/man4/man4.i386/cs.4 +++ b/share/man/man4/man4.i386/cs.4 @@ -47,7 +47,8 @@ driver provides support for ISA ethernet adapters based on the .Tn Crystal Semiconductor CS8900 and .Tn CS8920 -NICs. These devices are used on the +NICs. +These devices are used on the .Tn IBM EtherJet ISA adapters and in many embedded applications where the high integration, small size and low cost of the CS89x0 family compensate for their drawbacks. @@ -62,7 +63,8 @@ Other parameters specified in .Pa /boot/device.hints will be used if present; the card may be soft-configured so these may be any valid -value. Adapters based on the CS8920 normally offer PnP configuration and the driver +value. +Adapters based on the CS8920 normally offer PnP configuration and the driver will detect the .Tn IBM EtherJet and the @@ -70,25 +72,29 @@ and the adapters automatically. .Pp Note that the CS8900 is limited to 4 IRQ values; these are normally implemented -as 5, 10, 11 and 12. The CS8920 has no such limitation. +as 5, 10, 11 and 12. +The CS8920 has no such limitation. .Pp Memory-mapped and DMA operation are not supported at this time. .Sh DIAGNOSTICS .Bl -diag .It "cs%d: full/half duplex negotiation timeout" -The attempt to negotiate duplex settings with the hub timed out. This may +The attempt to negotiate duplex settings with the hub timed out. +This may indicate a cabling problem or a faulty or incompatible hub. .It "cs%d: failed to enable " The CS89x0 failed to select the nominated media, either because it is not present or not operating correctly. .It "cs%d: No EEPROM, assuming defaults" -The CS89x0 does not have an EEPROM, or the EEPROM is hopelessly damaged. Operation +The CS89x0 does not have an EEPROM, or the EEPROM is hopelessly damaged. +Operation will only be successful if the configuration entry lists suitable values for the adapter. .It "cs%d: Invalid irq" The IRQ specified in the configuration entry is not valid for the adapter. .It "cs%d: Could not allocate memory for NIC" -There is a critical memory shortage. The adapter will not function. +There is a critical memory shortage. +The adapter will not function. .It "cs%d: Adapter has no media" The adapter is not configured for a specific media type. The media type will have @@ -98,11 +104,13 @@ The PnP probe code found a recognised adapter, but the adapter is disabled. .It "failed to read pnp parms" A PnP adapter was found, but configuration parameters for it could not be read. .It "failed to pnp card parameters" -The parameters obtained via PnP were not accepted by the driver. The adapter +The parameters obtained via PnP were not accepted by the driver. +The adapter may not function. .El .Sh CAVEATS -The CS89x0 family of adapters have a very small RAM buffer (4K). This may +The CS89x0 family of adapters have a very small RAM buffer (4K). +This may cause problems with extremely high network loads or bursty network traffic. In particular, NFS operations should be limited to 1k read/write transactions in order to avoid overruns. diff --git a/share/man/man4/man4.i386/ct.4 b/share/man/man4/man4.i386/ct.4 index 85ee11bdd20b..c31877c13b66 100644 --- a/share/man/man4/man4.i386/ct.4 +++ b/share/man/man4/man4.i386/ct.4 @@ -123,9 +123,12 @@ host adapters. .Xr scsi 4 .Sh NOTES Historically, the driver for the Cronyx Tau WAN adapters was -ct. This device name was changed to ctau to avoid conflicts -with this pc98 ct driver. The network device name for ctau -is 'ct'. Please see +ct. +This device name was changed to ctau to avoid conflicts +with this pc98 ct driver. +The network device name for ctau +is 'ct'. +Please see .Xr ctau 4 for the details for that device. .Sh HISTORY diff --git a/share/man/man4/man4.i386/ex.4 b/share/man/man4/man4.i386/ex.4 index 5ece8b3e502f..038213b1a4d5 100644 --- a/share/man/man4/man4.i386/ex.4 +++ b/share/man/man4/man4.i386/ex.4 @@ -41,8 +41,10 @@ and Pro/10+ Ethernet cards based on the Intel i82595 chip. The Olicom OC2220 is also supported. .Pp The card will be searched for in the -I/O address range 0x200 - 0x3a0. If the IRQ will be -read from the EEPROM on the card. For correct operation on newer +I/O address range 0x200 - 0x3a0. +If the IRQ will be +read from the EEPROM on the card. +For correct operation on newer cards the Plug-N-Play support should be disabled. .Sh DIAGNOSTICS .Bl -diag diff --git a/share/man/man4/man4.i386/fe.4 b/share/man/man4/man4.i386/fe.4 index 5133d9162745..51f04ecf1c81 100644 --- a/share/man/man4/man4.i386/fe.4 +++ b/share/man/man4/man4.i386/fe.4 @@ -326,7 +326,8 @@ This manual page was written by This document and the associated software may be used, modified, copied, distributed, and sold, in both source and binary form provided that the above copyright, these terms and the following disclaimer are -retained. The name of the author and/or the contributor may not be +retained. +The name of the author and/or the contributor may not be used to endorse or promote products derived from this document and the associated software without specific prior written permission. .Pp @@ -335,7 +336,8 @@ AND THE CONTRIBUTOR .Dq AS IS AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR -PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR THE +PURPOSE ARE DISCLAIMED. +IN NO EVENT SHALL THE AUTHOR OR THE CONTRIBUTOR BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR diff --git a/share/man/man4/man4.i386/io.4 b/share/man/man4/man4.i386/io.4 index b1110d9f3822..a3cb91211d4e 100644 --- a/share/man/man4/man4.i386/io.4 +++ b/share/man/man4/man4.i386/io.4 @@ -44,7 +44,8 @@ Any process that holds a file descriptor on open will get its .Em IOPL bits in the flag register set, thus allowing it to perform direct -I/O operations. This can be useful in order to write userland +I/O operations. +This can be useful in order to write userland programs that handle some hardware directly. Note that even read-only access will grant the full I/O privileges. .Pp diff --git a/share/man/man4/man4.i386/mse.4 b/share/man/man4/man4.i386/mse.4 index a6dfa6914851..c3b07c8b146d 100644 --- a/share/man/man4/man4.i386/mse.4 +++ b/share/man/man4/man4.i386/mse.4 @@ -45,7 +45,8 @@ is usually 0x23c. Some cards may also be set to use the secondary port address at 0x238. The interface cards require a single IRQ, which may be -2, 3, 4 or 5. Some cards may offer additional IRQs. +2, 3, 4 or 5. +Some cards may offer additional IRQs. The port number and the IRQ number are configured by jumpers on the cards or by software provided with the card. .Pp @@ -55,7 +56,8 @@ some interface cards. It may be 15, 30, 60 or 120Hz. .Pp The difference between the two types of the mice is not in mouse devices -(in fact they are exactly the same). But in the circuit on the interface +(in fact they are exactly the same). +But in the circuit on the interface cards. This means that the device from a bus mouse package can be connected to the interface card from an InPort mouse package, or vice diff --git a/share/man/man4/man4.i386/ndis.4 b/share/man/man4/man4.i386/ndis.4 index 76b9c219eb13..18e55beec332 100644 --- a/share/man/man4/man4.i386/ndis.4 +++ b/share/man/man4/man4.i386/ndis.4 @@ -49,33 +49,41 @@ network drivers to be used with The .Nm driver is provided in source code form and must be combined with -the Windows(r) driver supplied with your network adapter. The +the Windows(r) driver supplied with your network adapter. +The .Nm driver uses the .Xr ndisapi 9 kernel subsystem to relocate and link the Windows(r) binary so -that it can be used in conjunction with native code. The +that it can be used in conjunction with native code. +The .Xr ndisapi 9 subsystem provides an interface between the NDIS API and the .Fx -networking infrastructure. The Windows(r) driver is essentially -fooled into thinking it's running on Windows(r). Note that this +networking infrastructure. +The Windows(r) driver is essentially +fooled into thinking it's running on Windows(r). +Note that this means the .Nm driver is only useful on x86 machines. .Pp To build a functional driver, the user must have a copy of the -driver distribution media for his or her card. From this distribution, +driver distribution media for his or her card. +From this distribution, the user must extract two files: the .SYS file containing the driver binary code, and its companion .INF file, which contains the definitions for driver-specific registry keys and other installation -data such as device identifiers. These two files can be converted +data such as device identifiers. +These two files can be converted into a .Pa ndis_driver_data.h file using the .Xr ndiscvt 8 -utility. This file contains a binary image of the driver plus -registry key data. When the +utility. +This file contains a binary image of the driver plus +registry key data. +When the .Nm driver loads, it will create .Xr sysctl 9 @@ -84,13 +92,17 @@ nodes for each registry key extracted from the .INF file. The .Nm driver is designed to support mainly ethernet and wireless -network devices with PCI and PCMCIA bus attachments. (Cardbus -devices are also supported as a subset of PCI.) It there can -support many different media types and speeds. One limitation +network devices with PCI and PCMCIA bus attachments. +(Cardbus +devices are also supported as a subset of PCI.) +It there can +support many different media types and speeds. +One limitation however, is that there is no consistent way to learn if an ethernet device is operating in full or half duplex mode. The NDIS API allows for a generic means for determining link -state and speed, but not the duplex setting. There may be +state and speed, but not the duplex setting. +There may be driver-specific registry keys to control the media setting which can be configured via the .Xr sysctl 8 diff --git a/share/man/man4/man4.i386/perfmon.4 b/share/man/man4/man4.i386/perfmon.4 index 27913825f7d3..f39ed07bae6b 100644 --- a/share/man/man4/man4.i386/perfmon.4 +++ b/share/man/man4/man4.i386/perfmon.4 @@ -46,7 +46,8 @@ capabilities of the .Tn Pentium and .Tn "Pentium Pro" -CPUs. These processors implement two internal counters which can be +CPUs. +These processors implement two internal counters which can be configured to measure a variety of events for either count or duration (in CPU cycles), as well as a cycle counter which counts clock cycles. The @@ -67,7 +68,8 @@ and processors. .Pp .Sy NOTA BENE : -The set of available events differs from processor to processor. It +The set of available events differs from processor to processor. +It is the responsibility of the programmer to ensure that the event numbers used are the correct ones for the CPU type being measured. .Pp @@ -119,22 +121,27 @@ returns the current configuration of the specified counter. .It Dv PMIOSTART .It Dv PMIOSTOP .Pq Li int -starts (stops) the specified counter. Due to hardware deficiencies, -counters must be started and stopped in numerical order. (That is to +starts (stops) the specified counter. +Due to hardware deficiencies, +counters must be started and stopped in numerical order. +(That is to say, counter 0 can never be stopped without first stopping counter 1.) The driver will .Em not enforce this restriction (since it may not be present in future CPUs). .It Dv PMIORESET .Pq Li int -reset the specified counter to zero. The counter should be stopped +reset the specified counter to zero. +The counter should be stopped with .Dv PMIOSTOP -before it is reset. All counters are automatically reset by +before it is reset. +All counters are automatically reset by .Dv PMIOSETUP . .It Dv PMIOREAD .Pq Li "struct pmc_data" -get the current value of the counter. The +get the current value of the counter. +The .Li pmc_data structure defines two fields: .Pp @@ -152,7 +159,8 @@ instruction on processors to read the counters directly. .It Dv PMIOTSTAMP .Pq Li "struct pmc_tstamp" -read the time stamp counter. The +read the time stamp counter. +The .Li pmc_tstamp structure defines two fields: .Pp @@ -166,7 +174,8 @@ the current value of the counter as a 64-bit integer It is important to note that the counter rate, as provided in the .Li pmct_rate field, is often incorrect because of calibration difficulties and -non-integral clock rates. This field should be considered more of a +non-integral clock rates. +This field should be considered more of a hint or sanity-check than an actual representation of the rate of clock ticks. .El diff --git a/share/man/man4/man4.i386/ray.4 b/share/man/man4/man4.i386/ray.4 index 8553d934edef..397dde469d37 100644 --- a/share/man/man4/man4.i386/ray.4 +++ b/share/man/man4/man4.i386/ray.4 @@ -253,7 +253,7 @@ Represents link corruption or non standard nodes in the network. The received .Tn IEEE 802.11 -data packet has a reserved (i.e. not allowed) subtype. +data packet has a reserved (i.e., not allowed) subtype. Represents link corruption or non standard nodes in the network. .It "ray?: MGT TODS/FROMDS wrong fc1 0x??" The received @@ -272,7 +272,7 @@ Benign, but might represent buggy firmware. The received .Tn IEEE 802.11 -management packet has a reserved (i.e. not allowed) +management packet has a reserved (i.e., not allowed) subtype. Represents link corruption or non standard nodes in the network. .It "ray?: open system authentication request" @@ -287,7 +287,7 @@ Self explanatory and for testing .Tn "Aviator Pro" interworking. .It "ray?: reserved authentication subtype 0x??" -An authentication request has been received for a reserved (i.e. not allowed) +An authentication request has been received for a reserved (i.e., not allowed) subtype. Represents link corruption or non standard nodes in the network. .It "ray?: CTL TODS/FROMDS wrong fc1 0x??" @@ -307,7 +307,7 @@ Benign, but might represent buggy firmware. The received .Tn IEEE 802.11 -control packet has a reserved (i.e. not allowed) +control packet has a reserved (i.e., not allowed) subtype. Represents link corruption or non standard nodes in the network. .It "ray?: bad ccs index 0x??" diff --git a/share/man/man4/man4.i386/scd.4 b/share/man/man4/man4.i386/scd.4 index ad8456dc9a76..bca2be9577fc 100644 --- a/share/man/man4/man4.i386/scd.4 +++ b/share/man/man4/man4.i386/scd.4 @@ -43,7 +43,8 @@ In The .Nm driver provides a data interface to the Sony CDU31 and CDU33A CD-ROM -drives. The drive must be hooked to a Sony proprietary interface +drives. +The drive must be hooked to a Sony proprietary interface card or a compatible clone. .Sh FILES .Bl -tag -width /dev/[r]scd0a -compact diff --git a/share/man/man4/man4.i386/spkr.4 b/share/man/man4/man4.i386/spkr.4 index 199a131899b6..e29ad6c18d52 100644 --- a/share/man/man4/man4.i386/spkr.4 +++ b/share/man/man4/man4.i386/spkr.4 @@ -68,7 +68,8 @@ a zero duration. .Pp The play-string language is modeled on the PLAY statement conventions of .Tn IBM -Advanced BASIC 2.0. The +Advanced BASIC 2.0. +The .Li MB , .Li MF , and @@ -80,7 +81,8 @@ feature and the slur mark are new. .Pp There are 84 accessible notes numbered 1-84 in 7 octaves, each running from C to B, numbered 0-6; the scale is equal-tempered A440 and octave 3 starts -with middle C. By default, the play function emits half-second notes with the +with middle C. +By default, the play function emits half-second notes with the last 1/16th second being `rest time'. .Pp Play strings are interpreted left to right as a series of play command groups; @@ -89,12 +91,15 @@ Play command groups are as follows: .Bl -tag -width CDEFGABxx .It Li CDEFGAB Letters A through G cause the corresponding note to be played in the -current octave. A note letter may optionally be followed by an +current octave. +A note letter may optionally be followed by an .Dq Em "accidental sign" , one of # + or -; the first two of these cause it to be sharped one -half-tone, the last causes it to be flatted one half-tone. It may +half-tone, the last causes it to be flatted one half-tone. +It may also be followed by a time value number and by sustain dots (see -below). Time values are interpreted as for the L command below. +below). +Time values are interpreted as for the L command below. .It Ns Li O Sy n If .Sy n @@ -109,8 +114,10 @@ When octave-tracking is on, interpretation of a pair of letter notes will change octaves if necessary in order to make the smallest possible jump between notes. Thus ``olbc'' will be played as -``olb>c'', and ``olcb'' as ``olc, < and O[0123456]. (The octave-locking +``olb>c'', and ``olcb'' as ``olc, < and O[0123456]. +(The octave-locking feature is not supported in .Tn IBM BASIC.) @@ -125,7 +132,8 @@ Play note being 1 to 84 or 0 for a rest of current time value. May be followed by sustain dots. .It Ns Li L Sy n -Sets the current time value for notes. The default is +Sets the current time value for notes. +The default is .Li L4 , quarter or crotchet notes. The lowest possible value is 1; values up @@ -142,10 +150,12 @@ Pause (rest), with interpreted as for .Li L Sy n . May be followed by -sustain dots. May also be written +sustain dots. +May also be written .Li ~ . .It Ns Li T Sy n -Sets the number of quarter notes per minute; default is 120. Musical +Sets the number of quarter notes per minute; default is 120. +Musical names for common tempi are: .Bd -literal -offset indent Tempo Beats Per Minute @@ -192,7 +202,8 @@ dotted twice, it is held 9/4, and three times would give 27/8. .Pp A note and its sustain dots may also be followed by a slur mark (underscore). This causes the normal micro-rest after the note to be filled in, slurring it -to the next one. (The slur feature is not supported in +to the next one. +(The slur feature is not supported in .Tn IBM BASIC.) .Pp @@ -206,9 +217,11 @@ There is no volume control. .Pp The action of two or more sustain dots does not reflect standard musical notation, in which each dot adds half the value of the previous dot -modifier, not half the value of the note as modified. Thus, a note dotted +modifier, not half the value of the note as modified. +Thus, a note dotted once is held for 3/2 of its undotted value; dotted twice, it is held 7/4, -and three times would give 15/8. The multiply-by-3/2 interpretation, +and three times would give 15/8. +The multiply-by-3/2 interpretation, however, is specified in the .Tn IBM BASIC manual and has been retained for diff --git a/share/man/man4/man4.i386/streams.4 b/share/man/man4/man4.i386/streams.4 index d872f9328495..84a572caa480 100644 --- a/share/man/man4/man4.i386/streams.4 +++ b/share/man/man4/man4.i386/streams.4 @@ -55,7 +55,7 @@ Hence, opening a stream device produces a result similar to what would be obtained by calling .Xr socket 2 . .Pp -Applications should never use this interface directly: STREAMS +Applications should never use this interface directly: STREAMS emulation is only provided as a service to support ABI requirements in the SVR4 environment which .Xr svr4 4 diff --git a/share/man/man4/man4.i386/vx.4 b/share/man/man4/man4.i386/vx.4 index ec430864eac7..4943026611eb 100644 --- a/share/man/man4/man4.i386/vx.4 +++ b/share/man/man4/man4.i386/vx.4 @@ -74,7 +74,8 @@ Use the UTP port. .Bl -diag .It "vx%d: not configured; kernel is built for only %d devices." There are not enough devices in the kernel configuration file for the number -of adapters present in the system. Add devices to the configuration file, +of adapters present in the system. +Add devices to the configuration file, rebuild the kernel, and reboot. .El .Pp @@ -82,7 +83,8 @@ All other diagnostics indicate either a hardware problem or a bug in the driver. .Sh CAVEATS Some early-revision 3c590 cards are defective and suffer from many receive -overruns, which cause lost packets. The author has attempted to implement +overruns, which cause lost packets. +The author has attempted to implement a test for it based on the information supplied by 3Com, but the test resulted mostly in spurious warnings. .Pp diff --git a/share/man/man4/man4.i386/wl.4 b/share/man/man4/man4.i386/wl.4 index 6dc31f4e1099..b6b7860182a8 100644 --- a/share/man/man4/man4.i386/wl.4 +++ b/share/man/man4/man4.i386/wl.4 @@ -41,28 +41,38 @@ The .Nm driver controls a radio lan card system made originally by -NCR, then ATT, now Lucent. The system is spread-spectrum radio -at around 915 MHz (or 2.4 GHz). With the supplied omni-directional antennae, +NCR, then ATT, now Lucent. +The system is spread-spectrum radio +at around 915 MHz (or 2.4 GHz). +With the supplied omni-directional antennae, about 400 feet (indoors, more outdoors) can be covered in circumference. -This card can talk to the companion (wlp0) pccard. Speeds vary +This card can talk to the companion (wlp0) pccard. +Speeds vary from 1 megabit to theoretically 2 megabits (roughly T1 in speed). .Pp The card has three fundamental hardware units, a so-called PSA or programmable storage area, a radio modem, -and a ethernet lan controller. The latter component is the +and a ethernet lan controller. +The latter component is the ancient (and not very honorable) Intel 82586 ethernet chip. Fundamentally it appears to the operating system as an ethernet system, -and speaks IEEE MAC addresses. The radio modem simply translates +and speaks IEEE MAC addresses. +The radio modem simply translates ethernet packets to/from radio packets, that are either at 2.4 GHz -or 915 MHz depending on the radio modem. It supports a collision -avoidance scheme. The lan controller +or 915 MHz depending on the radio modem. +It supports a collision +avoidance scheme. +The lan controller supports promiscuous mode, broadcast, and multicasting (although there is a glitch -in the latter). "It thinks it is ethernet". +in the latter). +"It thinks it is ethernet". .Pp How it is used -depends on the kind of antennae deployed with it. Point to point -applications are possible as are ethernet-like lan use. The vendor +depends on the kind of antennae deployed with it. +Point to point +applications are possible as are ethernet-like lan use. +The vendor ships an omni-directional antennae that works in the vicinity of 400 feet (indoors). Point to point antennae can be purchased that will go miles. @@ -72,7 +82,8 @@ Typically minimally an IRQ, port, and Network ID must be supplied. Michael Smith's .Xr wlconfig 8 utility can now be used to do this work from -the UNIX side. The card is "not" plug and play. +the UNIX side. +The card is "not" plug and play. The network id controls whether one set of cards can hear another. If different, cards will read physical packets, but they will be discarded by the radio modem. @@ -85,13 +96,17 @@ variables are as follows: .Bl -diag .It "machdep.wl_xmit_delay " This variable will cause the driver to insert a delay on transmit. -250 is the default. The delay should probably be a bit longer -on faster cpus and less on slower cpus. It exists because the 82586 +250 is the default. +The delay should probably be a bit longer +on faster cpus and less on slower cpus. +It exists because the 82586 was not designed to work with Pentium-speed cpu systems and if overdriven will have copious xmit side errors. .It machdep.wl_ignore_nwid <0 | 1> -This switch defaults to 0; i.e., the nwid is not ignored. It can -be set to 1 to cause the nwid to not be used. This may be useful +This switch defaults to 0; i.e., the nwid is not ignored. +It can +be set to 1 to cause the nwid to not be used. +This may be useful when the device is in promiscuous mode as one can watch for all packets and ignore nwid differences. .It machdep.wl_xmit_watch @@ -99,33 +114,42 @@ This switch is not currently useful. .It machdep.wl_gather_snr This switch is not currently useful. .Pp -There is also a signal strength cache in the driver. It may be interrogated +There is also a signal strength cache in the driver. +It may be interrogated with .Xr wlconfig 8 . Incoming packets are checked for certain hardware radio-modem values including signal strength, silence, and quality, which range fro 0..63, 0..63, and 0..15 -respectively. Thus one can read out signal strenth values to see -how close/far peer nodes are. The signal strength cache is indexed by +respectively. +Thus one can read out signal strenth values to see +how close/far peer nodes are. +The signal strength cache is indexed by sender MAC address. -There are two sysctls that change how it filters packets. Both are on +There are two sysctls that change how it filters packets. +Both are on by default. .It machdep.wl_wlcache_mcastonly <0 | 1> By default this switch is on. It forces the cache to filter out -unicast packets. Only broadcast or multicast packets are accepted. +unicast packets. +Only broadcast or multicast packets are accepted. .It machdep.wl_wlcache_iponly <0 | 1> -By default this switch is on. It forces the driver to discard non-IP -packets and also stores the IP src address. ARP packets are ignored, +By default this switch is on. +It forces the driver to discard non-IP +packets and also stores the IP src address. +ARP packets are ignored, as are any other network protocol barring IPv4 packets. .El .Sh CAVEATS -The 82586 has numerous defects. It may experience transmit-side +The 82586 has numerous defects. +It may experience transmit-side errors when modern faster cpus send packets at it faster than it can handle. The driver (and probably the chip) does not support an all multicast mode. As a result, it can be used with applications like .Xr mrouted 8 , -but it must go into promiscuous mode for that to work. The driver +but it must go into promiscuous mode for that to work. +The driver is slow to change modes from "normal" to promiscuous mode, presumably due to delays in the configuration code. .Sh SEE ALSO @@ -139,9 +163,11 @@ The driver was written by .An Anders Klemets (thousands of years ago?) and -appears to be based on an even older Intel 82586 driver. The 82586 +appears to be based on an even older Intel 82586 driver. +The 82586 controller was one of the first (if not the first?) integrated lan -controller on the block. That does not mean it was the best either. +controller on the block. +That does not mean it was the best either. Anders ported and or created a driver for the ISA wavelan and PCCARD wavelan system too (wlp). .An Robert T. Morris, Jr. @@ -150,12 +176,16 @@ ported the Mach drivers to BSDI. ported them to .Fx 2.1 . .An Michael Smith -ported the wl driver only to 2.2.2. Jim and Michael have been -maintaining them. The current state of the driver is NOT ANYONE'S -FAULT. Thanks to +ported the wl driver only to 2.2.2. +Jim and Michael have been +maintaining them. +The current state of the driver is NOT ANYONE'S +FAULT. +Thanks to .An Bernie Doehner and .An Robert Buaas for contributions. .Sh AUTHORS -Too numerous to mention. See above. +Too numerous to mention. +See above. -- cgit v1.2.3