aboutsummaryrefslogtreecommitdiff
path: root/libcuda
AgeCommit message (Collapse)Author
2018-10-11Merging the latest dev from the public branchtgrogers
2018-10-05Merge pull request #78 from fusiled/devTim Rogers
libcuda: removed comment from cuobjdump.l
2018-10-05Merge pull request #76 from tgrogers/devTim Rogers
Changes to allow correct occupancy with newer cards, while still running PTXPlus
2018-09-20libcuda: removed comment from cuobjdump.lMatteo Maria Fusi
This comment made the build fail in ubuntu 18.04 due to a flex error. This error is probably produced by a flex bug.
2018-06-30changing the flow s.t. you cannot compile the simulator with a different ↵tgrogers
version of CUDA from the app.
2018-06-29Fixing the PTXPLUS + new CUDA execution case.tgrogers
2018-06-29Allowing modern GPU configrations to properly run PTXPLUS.tgrogers
There are several problems with this: 1) To get proper occupancy, based on register file usage, you must querry ptxas with the sm version that corresponds to the register usage in the config file. To enable this, a new config option has been added that determines what sm version you pass to ptxas to compute reg-usage. This configuration option is always required in the gpgpusim.config file 2) If you are running PTXPLUS with a modern card (i.e. volta/pascal), you need ptxas from CUDA 9.1. However since PTXPLUS only supports sm_13 - you need a version of CUDA where cudaobjectdump supports sm_1x. This ended at CUDA 5 - and PTXPLUS requires CUDA 4.2. Therefore, to run PTXPLUS on a modern card, you need CUDA 4.2 + modern CUDA installed. To fascilitate this, a new envronment varaible is added and the setup envrionment script prints an appropraite warning if you are using a newer CUDA. We have tried to make this as fail-proof as possible - and die appropraitely when something is wrong.
2018-06-29Whitespace edits - removing tabs - replacing with spacestgrogers
2018-06-29Adding a print to indicate which PTX version we are actually usingtgrogers
2018-04-22Some classes were referred to as a class and a struct (reported as clang ↵Nathan Conrad
warnings). This makes these consistent.
2018-03-28(attempt to) fix compile error in last checkinTor Aamodt
2018-03-28fix compile errors on Ubuntu LTS 16.04Tor Aamodt
2018-03-26Don't directly pass strings to printf (beacuse they wouldn't be escapped). ↵Nathan Conrad
Clang gives warnings about this.
2018-03-25succeed if just remining the GPU that pinned memory can beusedtgrogers
2018-03-25Add lonestar tick supportMengchi Zhang
Signed-off-by: Mengchi Zhang <[email protected]>
2018-03-24Adding in a patch to get lonestargpu to worktgrogers
2017-05-09Merge branch 'dev' into devTim Rogers
2016-10-28support for CUDA 8.0 PTXAhmed ElTantawy
2016-09-05Merge pull request #28 from jwang323/cdp_cleangpgpu-sim
Initial support of CUDA Dynamic Parallelism on GPGPUSim
2016-07-06ADD: add knob to enable CDP in gpgpusim configJin Wang
2016-07-05BUG: PTX section id. ADD: cudaDeviceSetLimit. BUG: parameter addresses for ↵Jin Wang
child kernels in CDP. BUG: .weak .entry and .weak .global directives in ptx file. BUG: empty_protected() for stream manager causes deadlock, change to empty()
2016-07-05ADD: handle child kernel name in mov instruction. ADD: detect call ↵Jin Wang
cudaGetParameterBufferV2 and call cudaLaunchDeviceV2
2016-07-05MOD: Hack warning for cuda device runtime lib function registrationJin Wang
2016-07-05BUG: cuobjdump for sm_35 with CDPJin Wang
2016-07-05MOD: add support for cuda65 ptx headerJin Wang
2016-07-04fix a silly typoAhmed ElTantawy
2016-06-10Vastly improved runtime by relating symbol tables to file namessspenst
2016-06-09Merge branch 'dev' of https://github.com/sspenst/gpgpu-sim_distribution into devspeverel
2016-06-09Modified runtime API to support synchronization with the null stream.speverel
2016-06-09Forgot to add a return valuesspenst
2016-06-09Added a basic implementation of cudaStreamCreateWithFlagssspenst
2016-06-07Updated the PTX section merging to be able to deal with multiple identifierssspenst
2016-06-07The ptx parser now recognizes the NC option for ld.global, however this ↵sspenst
option is not actually implemented
2016-06-07Added support for cudaMemcpyDefault flag in cudaMemcpy. Also increased the ↵speverel
maximum allowable memory to 2GB and the compute version to 5.2.
2016-06-03Added mergeSectionList to combine any PTX files that remain after pruning ↵sspenst
into a single file.
2016-06-02Added handling of .cc option for arithmetic instructions. NOTE: Only made ↵speverel
changes to parse instructions. Carry functionality NOT fully implemented; .cc instructions function like their unmodified ueqivelents. Also modified GTX750Ti config to model L1 data cache as simply not being used for global loads (instead of not existing at all). Changed ptxinfo parsing to avoid crashing when info includes texture information.
2016-05-13changed the SM version flag for generating ptxinfo to be dynamic instead of 20sspenst
2016-05-13remove the hackish comment and assign a default identifier for CUDA>=7.5Ahmed ElTantawy
2016-05-13detailed error message when minimum found ptx capability is still larger ↵Ahmed ElTantawy
than the maximum forced capability
2016-05-13getting the parsing compatible with previous version - tested on 4.2, 5.5 ↵Ahmed ElTantawy
and 7.5
2016-05-12added an assertion to pruneSectionListsspenst
2016-05-12Commented out line that checks for identifier in PTX file so that sim runs ↵speverel
in 7.5
2016-05-12initial support for cuda 7.5 parsingElTantawy
2016-05-12Adding in changes to put generated code files in the build directory instead ↵tgrogers
of the source directory - since this causes issues for parallel buildwq
2015-03-04initial support for CUDA 5.0, 5.5, 6.0 to get template from SDK runningAhmed ElTantawy
2014-08-14 Changing the make flow so all the generated files, both object and code ↵Tim Rogers
generated go in a directory independent of the source directory [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 16503]
2014-08-14Removing a slew of code still compiled with gcc and the need for a bunch of ↵Tim Rogers
external C linkage [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 15912]
2014-08-14define the cudaFuncSetCacheConfig for CUDART_VERSION >= 3000Ahmed El-Shafiey
[git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 15818]
2014-08-14 - Adding support for cudaFuncSetCacheConfig API, that allows changing theAhmed El-Shafiey
L1 Cache and Shared Memory configurations across kernels. The support enable the user to specify two more configurations (Preferred L1) or (Preferred Shared Memory) besides the default config. If the cudaFuncSetCacheConfig API is used to set the cache configuration of a specific kernel to either of these configuration (cudaFuncCachePreferShared, cudaFuncCachePreferL1), the simulator will change the cache configuration at kernel launch accordingly, if there is no alternative configurations provided to the simulator it will use the default configurations with a warning message display [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 15816]
2014-08-14 Fixing a slew of compiler warningsTim Rogers
[git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 15218]
generated by cgit v1.3 (git 2.53.0) at 2026-07-13 20:49:15 +0000