1Accessing PCI device resources through sysfs 2-------------------------------------------- 3 4sysfs, usually mounted at /sys, provides access to PCI resources on platforms 5that support it. For example, a given bus might look like this: 6 7 /sys/devices/pci0000:17 8 |-- 0000:17:00.0 9 | |-- class 10 | |-- config 11 | |-- device 12 | |-- enable 13 | |-- irq 14 | |-- local_cpus 15 | |-- remove 16 | |-- resource 17 | |-- resource0 18 | |-- resource1 19 | |-- resource2 20 | |-- revision 21 | |-- rom 22 | |-- subsystem_device 23 | |-- subsystem_vendor 24 | `-- vendor 25 `-- ... 26 27The topmost element describes the PCI domain and bus number. In this case, 28the domain number is 0000 and the bus number is 17 (both values are in hex). 29This bus contains a single function device in slot 0. The domain and bus 30numbers are reproduced for convenience. Under the device directory are several 31files, each with their own function. 32 33 file function 34 ---- -------- 35 class PCI class (ascii, ro) 36 config PCI config space (binary, rw) 37 device PCI device (ascii, ro) 38 enable Whether the device is enabled (ascii, rw) 39 irq IRQ number (ascii, ro) 40 local_cpus nearby CPU mask (cpumask, ro) 41 remove remove device from kernel's list (ascii, wo) 42 resource PCI resource host addresses (ascii, ro) 43 resource0..N PCI resource N, if present (binary, mmap, rw[1]) 44 resource0_wc..N_wc PCI WC map resource N, if prefetchable (binary, mmap) 45 revision PCI revision (ascii, ro) 46 rom PCI ROM resource, if present (binary, ro) 47 subsystem_device PCI subsystem device (ascii, ro) 48 subsystem_vendor PCI subsystem vendor (ascii, ro) 49 vendor PCI vendor (ascii, ro) 50 51 ro - read only file 52 rw - file is readable and writable 53 wo - write only file 54 mmap - file is mmapable 55 ascii - file contains ascii text 56 binary - file contains binary data 57 cpumask - file contains a cpumask type 58 59[1] rw for RESOURCE_IO (I/O port) regions only 60 61The read only files are informational, writes to them will be ignored, with 62the exception of the 'rom' file. Writable files can be used to perform 63actions on the device (e.g. changing config space, detaching a device). 64mmapable files are available via an mmap of the file at offset 0 and can be 65used to do actual device programming from userspace. Note that some platforms 66don't support mmapping of certain resources, so be sure to check the return 67value from any attempted mmap. The most notable of these are I/O port 68resources, which also provide read/write access. 69 70The 'enable' file provides a counter that indicates how many times the device 71has been enabled. If the 'enable' file currently returns '4', and a '1' is 72echoed into it, it will then return '5'. Echoing a '0' into it will decrease 73the count. Even when it returns to 0, though, some of the initialisation 74may not be reversed. 75 76The 'rom' file is special in that it provides read-only access to the device's 77ROM file, if available. It's disabled by default, however, so applications 78should write the string "1" to the file to enable it before attempting a read 79call, and disable it following the access by writing "0" to the file. Note 80that the device must be enabled for a rom read to return data successfully. 81In the event a driver is not bound to the device, it can be enabled using the 82'enable' file, documented above. 83 84The 'remove' file is used to remove the PCI device, by writing a non-zero 85integer to the file. This does not involve any kind of hot-plug functionality, 86e.g. powering off the device. The device is removed from the kernel's list of 87PCI devices, the sysfs directory for it is removed, and the device will be 88removed from any drivers attached to it. Removal of PCI root buses is 89disallowed. 90 91Accessing legacy resources through sysfs 92---------------------------------------- 93 94Legacy I/O port and ISA memory resources are also provided in sysfs if the 95underlying platform supports them. They're located in the PCI class hierarchy, 96e.g. 97 98 /sys/class/pci_bus/0000:17/ 99 |-- bridge -> ../../../devices/pci0000:17 100 |-- cpuaffinity 101 |-- legacy_io 102 `-- legacy_mem 103 104The legacy_io file is a read/write file that can be used by applications to 105do legacy port I/O. The application should open the file, seek to the desired 106port (e.g. 0x3e8) and do a read or a write of 1, 2 or 4 bytes. The legacy_mem 107file should be mmapped with an offset corresponding to the memory offset 108desired, e.g. 0xa0000 for the VGA frame buffer. The application can then 109simply dereference the returned pointer (after checking for errors of course) 110to access legacy memory space. 111 112Supporting PCI access on new platforms 113-------------------------------------- 114 115In order to support PCI resource mapping as described above, Linux platform 116code should ideally define ARCH_GENERIC_PCI_MMAP_RESOURCE and use the generic 117implementation of that functionality. To support the historical interface of 118mmap() through files in /proc/bus/pci, platforms may also set HAVE_PCI_MMAP. 119 120Alternatively, platforms which set HAVE_PCI_MMAP may provide their own 121implementation of pci_mmap_page_range() instead of defining 122ARCH_GENERIC_PCI_MMAP_RESOURCE. 123 124Platforms which support write-combining maps of PCI resources must define 125arch_can_pci_mmap_wc() which shall evaluate to non-zero at runtime when 126write-combining is permitted. Platforms which support maps of I/O resources 127define arch_can_pci_mmap_io() similarly. 128 129Legacy resources are protected by the HAVE_PCI_LEGACY define. Platforms 130wishing to support legacy functionality should define it and provide 131pci_legacy_read, pci_legacy_write and pci_mmap_legacy_page_range functions. 132