User:ZyMOS/Howto configure the linux kernel/drivers/atm

User:ZyMOS/Howto configure the linux kernel / drivers / atm


 * ATM device configuration
 * ATM device configuration

ATM drivers

 * depends on NETDEVICES && ATM


 * Option: ATM_DUMMY
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "Dummy ATM driver"
 * depends on ATM
 * Dummy ATM driver. Useful for proxy signalling, testing, and development. If unsure, say N.


 * Option: ATM_TCP
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "ATM over TCP"
 * depends on INET && ATM
 * ATM over TCP driver. Useful mainly for development and for experiments. If unsure, say N.


 * Option: ATM_LANAI
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "Efficient Networks Speedstream 3010"
 * depends on PCI && ATM
 * Supports ATM cards based on the Efficient Networks "Lanai" chipset such as the Speedstream 3010 and the ENI-25p. The Speedstream 3060 is currently not supported since we don't have the code to drive the on-board Alcatel DSL chipset (yet).


 * Option: ATM_ENI
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "Efficient Networks ENI155P"
 * depends on PCI && ATM
 * Driver for the Efficient Networks ENI155p series and SMC ATM Power155 155 Mbps ATM adapters. Both, the versions with 512KB and 2MB on-board RAM (Efficient calls them "C" and "S", respectively), and the FPGA and the ASIC Tonga versions of the board are supported. The driver works with MMF (-MF or ...F) and UTP-5 (-U5 or ...D) adapters.
 * To compile this driver as a module, choose M here: the module will be called eni.


 * Option: ATM_ENI_DEBUG
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable extended debugging"
 * depends on ATM_ENI
 * Extended debugging records various events and displays that list when an inconsistency is detected. This mechanism is faster than generally using printks, but still has some impact on performance. Note that extended debugging may create certain race conditions itself. Enable this ONLY if you suspect problems with the driver.


 * Option: ATM_ENI_TUNE_BURST
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Fine-tune burst settings"
 * depends on ATM_ENI
 * In order to obtain good throughput, the ENI NIC can transfer multiple words of data per PCI bus access cycle. Such a multi-word transfer is called a burst.
 * The default settings for the burst sizes are suitable for most PCI chipsets. However, in some cases, large bursts may overrun buffers in the PCI chipset and cause data corruption. In such cases, large bursts must be disabled and only (slower) small bursts can be used. The burst sizes can be set independently in the send (TX) and receive (RX) direction.
 * Note that enabling many different burst sizes in the same direction may increase the cost of setting up a transfer such that the resulting throughput is lower than when using only the largest available burst size.
 * Also, sometimes larger bursts lead to lower throughput, e.g. on an Intel 440FX board, a drop from 135 Mbps to 103 Mbps was observed when going from 8W to 16W bursts.


 * Option: ATM_ENI_BURST_TX_16W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 16W TX bursts (discouraged)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst sixteen words at once in the send direction. This may work with recent PCI chipsets, but is known to fail with older chipsets.


 * Option: ATM_ENI_BURST_TX_8W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 8W TX bursts (recommended)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst eight words at once in the send direction. This is the default setting.


 * Option: ATM_ENI_BURST_TX_4W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 4W TX bursts (optional)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst four words at once in the send direction. You may want to try this if you have disabled 8W bursts. Enabling 4W if 8W is also set may or may not improve throughput.


 * Option: ATM_ENI_BURST_TX_2W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 2W TX bursts (optional)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst two words at once in the send direction. You may want to try this if you have disabled 4W and 8W bursts. Enabling 2W if 4W or 8W are also set may or may not improve throughput.


 * Option: ATM_ENI_BURST_RX_16W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 16W RX bursts (discouraged)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst sixteen words at once in the receive direction. This may work with recent PCI chipsets, but is known to fail with older chipsets.


 * Option: ATM_ENI_BURST_RX_8W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 8W RX bursts (discouraged)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst eight words at once in the receive direction. This may work with recent PCI chipsets, but is known to fail with older chipsets, such as the Intel Neptune series.


 * Option: ATM_ENI_BURST_RX_4W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 4W RX bursts (recommended)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst four words at once in the receive direction. This is the default setting. Enabling 4W if 8W is also set may or may not improve throughput.


 * Option: ATM_ENI_BURST_RX_2W
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable 2W RX bursts (optional)"
 * depends on ATM_ENI_TUNE_BURST
 * Burst two words at once in the receive direction. You may want to try this if you have disabled 4W and 8W bursts. Enabling 2W if 4W or 8W are also set may or may not improve throughput.


 * Option: ATM_FIRESTREAM
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "Fujitsu FireStream (FS50/FS155) "
 * depends on PCI && ATM
 * Driver for the Fujitsu FireStream 155 (MB86697) and FireStream 50 (MB86695) ATM PCI chips.
 * To compile this driver as a module, choose M here: the module will be called firestream.


 * Option: ATM_ZATM
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "ZeitNet ZN1221/ZN1225"
 * depends on PCI && ATM
 * Driver for the ZeitNet ZN1221 (MMF) and ZN1225 (UTP-5) 155 Mbps ATM adapters.
 * To compile this driver as a module, choose M here: the module will be called zatm.


 * Option: ATM_ZATM_DEBUG
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable extended debugging"
 * depends on ATM_ZATM
 * Extended debugging records various events and displays that list when an inconsistency is detected. This mechanism is faster than generally using printks, but still has some impact on performance. Note that extended debugging may create certain race conditions itself. Enable this ONLY if you suspect problems with the driver.


 * bool 'Rolfs TI TNETA1570' CONFIG_ATM_TNETA1570 y
 * if [ "$CONFIG_ATM_TNETA1570" = "y" ]; then
 * bool ' Enable extended debugging' CONFIG_ATM_TNETA1570_DEBUG n
 * fi


 * Option: ATM_NICSTAR
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "IDT 77201 (NICStAR) (ForeRunnerLE)"
 * depends on PCI && ATM && !64BIT
 * The NICStAR chipset family is used in a large number of ATM NICs for 25 and for 155 Mbps, including IDT cards and the Fore ForeRunnerLE series. Say Y if you have one of those.
 * To compile this driver as a module, choose M here: the module will be called nicstar.


 * Option: ATM_NICSTAR_USE_SUNI
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Use suni PHY driver (155Mbps)"
 * depends on ATM_NICSTAR
 * Support for the S-UNI and compatible PHYsical layer chips. These are found in most 155Mbps NICStAR based ATM cards, namely in the ForeRunner LE155 cards. This driver provides detection of cable~ removal and reinsertion and provides some statistics. This driver doesn't have removal capability when compiled as a module, so if you need that capability don't include S-UNI support (it's not needed to make the card work).


 * Option: ATM_NICSTAR_USE_IDT77105
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Use IDT77015 PHY driver (25Mbps)"
 * depends on ATM_NICSTAR
 * Support for the PHYsical layer chip in ForeRunner LE25 cards. In addition to cable removal/reinsertion detection, this driver allows you to control the loopback mode of the chip via a dedicated IOCTL. This driver is required for proper handling of temporary carrier loss, so if you have a 25Mbps NICStAR based ATM card you must say Y.


 * Option: ATM_IDT77252
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "IDT 77252 (NICStAR II)"
 * depends on PCI && ATM
 * Driver for the IDT 77252 ATM PCI chips.
 * To compile this driver as a module, choose M here: the module will be called idt77252.


 * Option: ATM_IDT77252_DEBUG
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable debugging messages"
 * depends on ATM_IDT77252
 * Somewhat useful debugging messages are available. The choice of messages is controlled by a bitmap. This may be specified as a module argument.  See the file  for the meanings of the bits in the mask.
 * When active, these messages can have a significant impact on the speed of the driver, and the size of your syslog files! When inactive, they will have only a modest impact on performance.


 * Option: ATM_IDT77252_RCV_ALL
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Receive ALL cells in raw queue"
 * depends on ATM_IDT77252
 * Enable receiving of all cells on the ATM link, that do not match an open connection in the raw cell queue of the driver. Useful for debugging or special applications only, so the safe answer is N.


 * Option: ATM_IDT77252_USE_SUNI
 * Kernel Versions: 2.6.15.6 ...
 * (on/off)
 * depends on ATM_IDT77252
 * default y


 * Option: ATM_AMBASSADOR
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "Madge Ambassador (Collage PCI 155 Server)"
 * depends on PCI && ATM
 * This is a driver for ATMizer based ATM card produced by Madge Networks Ltd. Say Y (or M to compile as a module named ambassador) here if you have one of these cards.


 * Option: ATM_AMBASSADOR_DEBUG
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable debugging messages"
 * depends on ATM_AMBASSADOR
 * Somewhat useful debugging messages are available. The choice of messages is controlled by a bitmap. This may be specified as a module argument (kernel command line argument as well?), changed dynamically using an ioctl (not yet) or changed by sending the string "Dxxxx" to VCI 1023 (where x is a hex digit).  See the file  for the meanings of the bits in the mask.
 * When active, these messages can have a significant impact on the speed of the driver, and the size of your syslog files! When inactive, they will have only a modest impact on performance.


 * Option: ATM_HORIZON
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "Madge Horizon [Ultra] (Collage PCI 25 and Collage PCI 155 Client)"
 * depends on PCI && ATM
 * This is a driver for the Horizon chipset ATM adapter cards once produced by Madge Networks Ltd. Say Y (or M to compile as a module named horizon) here if you have one of these cards.


 * Option: ATM_HORIZON_DEBUG
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable debugging messages"
 * depends on ATM_HORIZON
 * Somewhat useful debugging messages are available. The choice of messages is controlled by a bitmap. This may be specified as a module argument (kernel command line argument as well?), changed dynamically using an ioctl (not yet) or changed by sending the string "Dxxxx" to VCI 1023 (where x is a hex digit).  See the file  for the meanings of the bits in the mask.
 * When active, these messages can have a significant impact on the speed of the driver, and the size of your syslog files! When inactive, they will have only a modest impact on performance.


 * Option: ATM_IA
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "Interphase ATM PCI x575/x525/x531"
 * depends on PCI && ATM && !64BIT
 * This is a driver for the Interphase (i)ChipSAR adapter cards which include a variety of variants in term of the size of the control memory (128K-1KVC, 512K-4KVC), the size of the packet memory (128K, 512K, 1M), and the PHY type (Single/Multi mode OC3, UTP155, UTP25, DS3 and E3). Go to:  for more info about the cards. Say Y (or M to compile as a module named iphase) here if you have one of these cards.
 * See the file  for further details.


 * Option: ATM_IA_DEBUG
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Enable debugging messages"
 * depends on ATM_IA
 * Somewhat useful debugging messages are available. The choice of messages is controlled by a bitmap. This may be specified as a module argument (kernel command line argument as well?), changed dynamically using an ioctl (Get the debug utility, iadbg, from ).
 * See the file  for the meanings of the bits in the mask.
 * When active, these messages can have a significant impact on the speed of the driver, and the size of your syslog files! When inactive, they will have only a modest impact on performance.


 * Option: ATM_FORE200E_MAYBE
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "FORE Systems 200E-series"
 * depends on (PCI || SBUS) && ATM
 * This is a driver for the FORE Systems 200E-series ATM adapter cards. It simultaneously supports PCA-200E and SBA-200E models on PCI and SBUS hosts. Say Y (or M to compile as a module named fore_200e) here if you have one of these ATM adapters.
 * Note that the driver will actually be compiled only if you additionally enable the support for PCA-200E and/or SBA-200E cards.
 * See the file  for further details.


 * Option: ATM_FORE200E_PCA
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "PCA-200E support"
 * depends on ATM_FORE200E_MAYBE && PCI
 * Say Y here if you want your PCA-200E cards to be probed.


 * Option: ATM_FORE200E_PCA_DEFAULT_FW
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Use default PCA-200E firmware (normally enabled)"
 * depends on ATM_FORE200E_PCA
 * Use the default PCA-200E firmware data shipped with the driver.
 * Normal users do not have to deal with the firmware stuff, so they should say Y here.

"Pathname of user-supplied binary firmware"
 * Option: ATM_FORE200E_PCA_FW
 * Kernel Versions: 2.6.15.6 ...
 * depends on ATM_FORE200E_PCA && !ATM_FORE200E_PCA_DEFAULT_FW
 * default ""
 * This defines the pathname of an alternative PCA-200E binary firmware image supplied by the user. This pathname may be absolute or relative to the drivers/atm directory.
 * The driver comes with an adequate firmware image, so normal users do not have to supply an alternative one. They just say Y to "Use default PCA-200E firmware" instead.


 * Option: ATM_FORE200E_SBA
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "SBA-200E support"
 * depends on ATM_FORE200E_MAYBE && SBUS
 * Say Y here if you want your SBA-200E cards to be probed.


 * Option: ATM_FORE200E_SBA_DEFAULT_FW
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Use default SBA-200E firmware (normally enabled)"
 * depends on ATM_FORE200E_SBA
 * Use the default SBA-200E firmware data shipped with the driver.
 * Normal users do not have to deal with the firmware stuff, so they should say Y here.

"Pathname of user-supplied binary firmware"
 * Option: ATM_FORE200E_SBA_FW
 * Kernel Versions: 2.6.15.6 ...
 * depends on ATM_FORE200E_SBA && !ATM_FORE200E_SBA_DEFAULT_FW
 * default ""
 * This defines the pathname of an alternative SBA-200E binary firmware image supplied by the user. This pathname may be absolute or relative to the drivers/atm directory.
 * The driver comes with an adequate firmware image, so normal users do not have to supply an alternative one. They just say Y to "Use default SBA-200E firmware", above.


 * Option: ATM_FORE200E_USE_TASKLET
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Defer interrupt work to a tasklet"
 * depends on (PCI || SBUS) && (ATM_FORE200E_PCA || ATM_FORE200E_SBA)
 * default n
 * This defers work to be done by the interrupt handler to a tasklet instead of hanlding everything at interrupt time. This may improve the responsive of the host.

"Maximum number of tx retries"
 * Option: ATM_FORE200E_TX_RETRY
 * Kernel Versions: 2.6.15.6 ...
 * depends on (PCI || SBUS) && (ATM_FORE200E_PCA || ATM_FORE200E_SBA)
 * default "16"
 * Specifies the number of times the driver attempts to transmit a message before giving up, if the transmit queue of the ATM card is transiently saturated.
 * Saturation of the transmit queue may occur only under extreme conditions, e.g. when a fast host continuously submits very small frames (<64 bytes) or raw AAL0 cells (48 bytes) to the ATM adapter.
 * Note that under common conditions, it is unlikely that you encounter a saturation of the transmit queue, so the retry mechanism never comes into play.

"Debugging level (0-3)"
 * Option: ATM_FORE200E_DEBUG
 * Kernel Versions: 2.6.15.6 ...
 * depends on (PCI || SBUS) && (ATM_FORE200E_PCA || ATM_FORE200E_SBA)
 * default "0"
 * Specifies the level of debugging messages issued by the driver. The verbosity of the driver increases with the value of this parameter.
 * When active, these messages can have a significant impact on the performances of the driver, and the size of your syslog files! Keep the debugging level to 0 during normal operations.


 * Option: ATM_FORE200E
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module)
 * depends on (PCI || SBUS) && (ATM_FORE200E_PCA || ATM_FORE200E_SBA)
 * default m if ATM_FORE200E_MAYBE!=y
 * default y if ATM_FORE200E_MAYBE=y


 * Option: ATM_HE
 * Kernel Versions: 2.6.15.6 ...
 * (on/off/module) "ForeRunner HE Series"
 * depends on PCI && ATM
 * This is a driver for the Marconi ForeRunner HE-series ATM adapter cards. It simultaneously supports the 155 and 622 versions.


 * Option: ATM_HE_USE_SUNI
 * Kernel Versions: 2.6.15.6 ...
 * (on/off) "Use S/UNI PHY driver"
 * depends on ATM_HE
 * Support for the S/UNI-Ultra and S/UNI-622 found in the ForeRunner HE cards. This driver provides carrier detection some statistics.