Skip to main content

What is Scatter-Gather DMA (Direct Memory Access)?


When a block move operation occurs, traditional DMA controllers need a contiguous (nonsegmented) block of physical memory. On most systems, it is difficult to get nonsegmented memory returned from the operating system.

DMA controllers can take segmented memory and perform a move operation on this memory using "scatter-gather" DMA. In this method, the DMA controllers store the starting addresses of all the memory segments. After the move operation starts, the DMA controller automatically feeds the DMA controller the start address of the next segment after the previous segment of memory is completed. Using this method, you do not need a contiguous block of memory from the operating system. This makes DMA ideal for non deterministic operating systems.

Comments

Popular posts from this blog

HMC Commands

Scripting directly within the HMC restricted shell - It can be done! for system in `lssyscfg -r sys -F " name,state" | sort | grep " ,Operating" | sed 's/,Operating//'` ; do     echo $system     echo "    LPAR            CPU    VCPU   MEM    OS"     for lpar in `lssyscfg -m $system -r lpar -F " name" | sort` ; do         defa ult_ prof = `ls sysc fg -r lpar -m $system --filter " lpar_names=$lpar" -F default_profile`         procs = `lssyscfg -r prof -m $system --filter " prof ile_ name s=$d efau lt_p rof, lpar _nam es=$ lpar " -F desired_proc_units`         vcpu = `lssyscfg -r prof -m $system --filter " prof ile_ name s=$d efau lt_p rof, lpar _nam es=$ lpar " -F desired_procs`     ...
Buffer to Buffer credits Define the maximum amount of data that can be sent prior to an acknowledgement Buffer credits are physical ASIC port or card memory resources and are finite in number as a function of cost Within a fabric, each port may have a different number of buffer credits The number of available buffer credits is communicated at fabric logon (FLOGI) One buffer credit allows a device to send one 2112 byte frame of data Before any data frames are sent, the transmitter sets a counter equal to the BB-credit value. For each data frame sent by the transmitter, the counter is decremented by one. Upon receipt of a data frame, the receiver sends a status frame (R_RDY) to the transmitter indicating that the data frame was received AND the buffer can receive another data frame. For each R_RDY received by the transmitter, the counter is incremented by one. BB_Credits are the “admission control” mechanism in FC to ensure that FC switches don’t run out of buffers ...

(SR-IOV) - Single Root I/O Virtualization (SR-IOV)

http://blog.scottlowe.org/2009/12/02/what-is-sr-iov/