aboutsummaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2020-09-23add missing headerThanos Makatos1-0/+1
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-09-23FIXME move vfio_user.h to correct locationThanos Makatos1-0/+0
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-09-23include vfio.h in vfio_user.hThanos Makatos1-0/+2
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-09-23add vfio_user.h to MakefileThanos Makatos2-0/+2
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-09-23fix attribute keywordThanos Makatos1-3/+3
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-09-23Add vfio_user.h from https://github.com/tmakatos/muser/pull/6Thanos Makatos1-0/+90
Signed-off-by: Changpeng Liu changpeng.liu@intel.com
2020-09-23Handle cleanup in init_socket for error outSwapnil Ingle1-35/+39
This patch adds cleanup during error out in init_socket. Also no need to create control socket path, bind() on AF_UNIX creates the path internally. Signed-off-by: Swapnil Ingle <swapnil.ingle@nutanix.com>
2020-09-17use correct socket fd in fcntlThanos Makatos1-1/+1
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com> Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-09-17don't reuse listening socket fd for connected socket fdThanos Makatos1-11/+12
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com> Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
2020-08-12Merge pull request #22 from franciozzy/cleanup_3Felipe Franciosi8-85/+89
Cleanup 3
2020-08-12samples/test_dma_map: Style fixesFelipe Franciosi1-2/+2
Fixes the type used in a format string and remove some whitespaces. Signed-off-by: Felipe Franciosi <felipe@nutanix.com> Signed-off-by: Swapnil Ingle <swapnil.ingle@nutanix.com>
2020-08-12muser.ko: Update for Linux 5.8Felipe Franciosi3-63/+24
This fixes vfio.diff to apply cleanly on Linux 5.8 and renames vm_insert_pages to _vm_insert_pages because an upstream function of the same name has been added to the kernel since. Future work will include replacing our version of vm_insert_pages with the upstream and redo all necessary testing. We also update travis to use v5.8. Signed-off-by: Felipe Franciosi <felipe@nutanix.com> Signed-off-by: Swapnil Ingle <swapnil.ingle@nutanix.com>
2020-08-12Split compilation unit of muser shared libraryFelipe Franciosi1-8/+19
This creates several smaller libraries out of the compilation units of the muser userspace library. It then links them together to build the shared library. In this process, we also generate unit test version of the libraries which include --coverage in the compilation and linking of those files. Those will be statically linked to unit tests later on. Signed-off-by: Felipe Franciosi <felipe@nutanix.com>
2020-08-12Bounce pci_info with pci_info_bounce()Felipe Franciosi1-1/+23
During lm_ctx_create, the pci_info struct from dev_info was being completely copied with a big memcpy(). That had the side-effect of copying pointers that were not allocated by us. Shortly after, copy_sparse_mmap_areas() rewrites these pointers with memory allocated by the library. As a result, any failure half-way through setting up the context would result in free_sparse_mmap_areas() calling free() on pointers we did not allocate (which resulted in errors or double frees). This fixes it by replacing the memcpy() with pci_info_bounce() which selectively copies only what is needed. Signed-off-by: Felipe Franciosi <felipe@nutanix.com>
2020-08-12sparse_mmap: Free bounced areas on lm_ctx_destroyFelipe Franciosi1-9/+10
We were leaking the areas bounced by copy_sparse_mmap_areas(). This fixes the leak by freeing the areas on the destructor. Signed-off-by: Felipe Franciosi <felipe@nutanix.com>
2020-08-12dma: Cope with dma being NULL in destructorFelipe Franciosi1-0/+4
On certain error paths, @dma will be NULL. This makes the destructor cope with it. Signed-off-by: Felipe Franciosi <felipe@nutanix.com>
2020-08-12Rename libmuser*c source filesFelipe Franciosi3-4/+9
For clarity and better project organisation towards unit tests, this renames libmuser.c to muser_ctx.c and libmuser_pci.c to muser_pci.c. CMakeLists is fixed accordingly including some other minor cleanup. Future work should further split muser_ctx.c so that unit tests can genuinely be written towards testing compilation units. Signed-off-by: Felipe Franciosi <felipe@nutanix.com>
2020-07-22clarify use of lm_irq_triggerThanos Makatos2-8/+8
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-22muser: use the correct PCI_COMMAND interrupt disable(ID) bitChangpeng Liu2-15/+2
When PCI_COMMAND interrupt disable bit set to 1, it means disable the pin-based INTx interrupt. For NVMe controller, it will use INTx first, and then switch to MSIX, the interrupt vector 0 is both valid for INTx and MSIX, so here we remove the check when posting an interrupt, just check the interrupt fd is valid or not. Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
2020-07-22add optional callback to notify context when DMA region is addedThanos Makatos2-0/+11
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-21document current work regarding VFIO-over-socketThanos Makatos1-0/+11
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-21focus README on VFIO-over-socketThanos Makatos1-81/+71
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-21change libvfio logging path to something more sensibleThanos Makatos1-1/+1
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-21make unmap_dma callback optionalThanos Makatos2-10/+34
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-20update gpio's command lineThanos Makatos1-1/+1
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-17update libpathtrapThanos Makatos1-0/+0
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-17refactor PCI capabilitiesThanos Makatos4-204/+140
This patch greatly refactors PCI capabilities. The main change is that we don't keep them in a list but simply store them in the PCI configuration space. Some corner cases aren't handled yet. It also fixes a bug where a capability isn't properly read depending on where it's located. Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-16add common definition for PCI capability headerThanos Makatos5-26/+58
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-15libpathtrap: fix compilation for gcc 10 / ld 2.34Thanos Makatos1-0/+0
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-15use HTTPS for libpathtrapThanos Makatos1-1/+1
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-15fix verbose loggingThanos Makatos4-53/+42
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-15handle PCI capabilities internallyThanos Makatos4-44/+344
Very few capabilities are handled, and implementation isn't great. Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-09minor improvements in non-blocking receive pathThanos Makatos1-15/+13
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-09Use MSG_DONTWAIT to receive requests if LM_FLAG_ATTACH_NB is specifiedBen Walker1-1/+18
If the user wanted non-blocking operation, don't block on the socket waiting for requests. Instead, use MSG_DONTWAIT to attempt to get the next request and return. The other operations all remain as blocking calls because they are not expected to ever need to wait. Signed-off-by: Ben Walker <benjamin.walker@intel.com>
2020-07-09All transport operations now take an lm_ctx_t as an argumentBen Walker1-23/+36
This is both for consistency and because the socket transport is going to need some place to cache things to handle partial sends and receives. Signed-off-by: Ben Walker <benjamin.walker@intel.com>
2020-07-08fix assertionThanos Makatos1-1/+1
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-08dma: notify device of DMA removal eventsThanos Makatos4-39/+106
This patch introduces a new device callback that is called when a DMA region is removed. The device must provide this callback and when the callback returns all references to the DMA regions must have been removed. This new functionality is necessary since a DMA region can be removed at any time, so the device must be given a chance to drop all references and perform the necessary cleanup. This patch also stores lm_ctx_t inside dma_controller_t so that we don't have to pass lm_ctx_t when calling dma_XXX functions. Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-08null device: add muser prefix in log callbackThanos Makatos1-1/+1
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-08Add lm_ctx_poll() as an alternative to lm_ctx_drive()Ben Walker2-53/+82
Using lm_ctx_poll() simply polls for work and returns instead of blocking. It is only valid when used on an lm_ctx created with LM_FLAG_ATTACH_NB. Signed-off-by: Ben Walker <benjamin.walker@intel.com>
2020-07-08Drop __attribute__((packed))from struct lm_ctxBen Walker1-13/+4
This makes it unsafe to take pointers to addresses within this data structure, and lots of the code does that. Signed-off-by: Ben Walker <benjamin.walker@intel.com>
2020-07-06pass log level to log functionThanos Makatos4-4/+4
This is useful when using another log function from the log function callback. Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-06don't leak caps when destroying LM contextThanos Makatos1-0/+1
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-03ignore VGA palette snoopingThanos Makatos1-0/+5
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-07-03handle memory write and invalidateThanos Makatos1-0/+13
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-06-24don't blindly trust region_hintThanos Makatos2-1/+6
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-06-23improve message when VFIO client closes connectionThanos Makatos1-1/+5
Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-06-12muser: use the correct PCI_COMMAND interrupt disable(ID) bitChangpeng Liu2-15/+2
When PCI_COMMAND interrupt disable bit set to 1, it means disable the pin-based INTx interrupt. For NVMe controller, it will use INTx first, and then switch to MSIX, the interrupt vector 0 is both valid for INTx and MSIX, so here we remove the check when posting an interrupt, just check the interrupt fd is valid or not. Signed-off-by: Changpeng Liu <changpeng.liu@intel.com> (cherry picked from commit 8ab7ca7d04365a7c39cd12ea45f8b9f7fa0afd47)
2020-06-04work in progress to make attaching to the transport optionally non-blockingThanos Makatos2-24/+109
This is required for using multiple controllers in SPDK. Signed-off-by: Thanos Makatos <thanos.makatos@nutanix.com>
2020-06-01let the control stack create IOMMU group symlink under /de/vfioThanos Makatos1-6/+0
2020-05-26FIXME loggingThanos Makatos1-1/+6
generated by cgit v1.1 at 2025-09-22 19:11:12 +0000