summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* qsort enough implementation, and corrected a little unwanted propagating ↵zilio nicolas2015-09-041-31/+17
| | | | description
* added cleanupzilio nicolas2015-09-042-8/+17
|
* almost finished regarding suren remarkszilio nicolas2015-09-0418-22/+47
|\
| * Fixes out-of-source builds and minor build issuesSuren A. Chilingaryan2015-09-0117-20/+42
| |
* | almost finished regarding suren remarkszilio nicolas2015-09-046-908/+1527
| |
* | some adressing regarding possible memory leakszilio nicolas2015-08-271-6/+22
| |
* | some cleaningzilio nicolas2015-08-272-71/+4
| |
* | xml files get by model and remove of validation modezilio nicolas2015-08-274-119/+132
| |
* | better public API and no more use of locals within pcilib_openzilio nicolas2015-08-273-186/+77
| |
* | put xml nodes pointers for banks and registers in pcilib_t, compil okzilio nicolas2015-08-278-113/+123
| |
* | pull from serverzilio nicolas2015-08-2626-154/+1152
|\|
| * Keep frame pointersSuren A. Chilingaryan2015-08-212-1/+6
| |
| * Add support for kernel 4.1Suren A. Chilingaryan2015-08-183-7/+7
| |
| * Improve bad events countingSuren A. Chilingaryan2015-08-141-21/+34
| |
| * Do not fail if PCI configuration is not fully available to unprivileged userSuren A. Chilingaryan2015-08-122-3/+9
| |
| * Fix support of older systems: Remove C11 derective and add rt library to the ↵Suren A. Chilingaryan2015-08-123-4/+6
| | | | | | | | link list
| * Disable STREAMING_CHECKS for better performanceSuren A. Chilingaryan2015-08-063-12/+8
| |
| * Fix compilation of the driverSuren A. Chilingaryan2015-08-062-2/+1
| |
| * In IPEDMA streaming mode put aside a single empty buffer to distinguish ↵Suren A. Chilingaryan2015-08-061-3/+14
| | | | | | | | between completely empty and full states of kernel ring buffer
| * Detect if IPEDMA operates in streaming modeSuren A. Chilingaryan2015-08-063-52/+32
| |
| * Protect access to the DMA engine with locksSuren A. Chilingaryan2015-08-063-5/+68
| |
| * Fix handling of inconsistent mutexesSuren A. Chilingaryan2015-08-061-3/+11
| |
| * In case of problematic locks report the errorSuren A. Chilingaryan2015-08-061-5/+5
| |
| * Provide pcilib_try_lock callSuren A. Chilingaryan2015-08-062-0/+11
| |
| * Minor fix for software registersSuren A. Chilingaryan2015-08-061-2/+4
| |
| * Install locking-related headersSuren A. Chilingaryan2015-08-051-1/+1
| |
| * Read model from environmental variable if definedSuren A. Chilingaryan2015-08-053-3/+8
| |
| * Use global locks to protect kmem allocation to prevent race while allocating ↵Suren A. Chilingaryan2015-08-055-14/+14
| | | | | | | | simmultaneously locking kmem pages and any other type of kmem
| * Protect mmaping BARs as wellSuren A. Chilingaryan2015-08-051-4/+13
| |
| * Protect with locks the initialization of software registersSuren A. Chilingaryan2015-08-051-5/+34
| |
| * Fix error-checking typo in kmemSuren A. Chilingaryan2015-08-051-2/+2
| |
| * Use bank addresses as kernel memory subtypes of software registers instead ↵Suren A. Chilingaryan2015-08-051-1/+1
| | | | | | | | of indexes in array
| * Integrate locking subsystem from Nicolas ZilioSuren A. Chilingaryan2015-08-0512-21/+873
| |\
| | * last modification+comments updatezilio nicolas2015-07-105-32/+49
| | |
| | * modified for last remarkszilio nicolas2015-07-072-20/+24
| | |
| | * clean version for lockszilio nicolas2015-07-0711-17/+346
| | |
| * | Make pcilib_streaming_action_t publicSuren A. Chilingaryan2015-07-243-11/+12
| | |
| * | Use 64-bit addressing in IPEDMA only for gen3 boards or if enforcedSuren A. Chilingaryan2015-07-243-10/+16
| |/
* | added READMEzilio nicolas2015-07-151-0/+23
| |
* | some modifzilio nicolas2015-07-076-9/+34
| |
* | modified pci to put banks before registerszilio nicolas2015-07-031-5/+5
| |
* | modified version to include future functionszilio nicolas2015-07-033-9/+372
| |
* | registers and banks support in xml v1. pci -ll works fine, but got segfault ↵zilio nicolas2015-07-0111-6/+2115
|/ | | | on pci -r name and pci -r name gives 0 always. might be due to the order in pci.c ------> ask suren
* Keep C++ compilers happySuren A. Chilingaryan2015-06-2214-1/+116
|
* Execute all operations on kernel_memory using plain pci model to avoid extra ↵Suren A. Chilingaryan2015-06-192-1/+5
| | | | references
* Integration of software registersSuren A. Chilingaryan2015-06-1910-6/+198
|\
| * new version of software_registerszilio nicolas2015-06-184-20/+47
| |
| * ready to run testszilio nicolas2015-06-151-0/+1
| |
| * some modificationszilio nicolas2015-06-125-8/+17
| |
| * added soft_regzilio nicolas2015-06-123-2/+150
| |