aboutsummaryrefslogtreecommitdiffstats
path: root/Documentation
diff options
context:
space:
mode:
Diffstat (limited to 'Documentation')
-rw-r--r--Documentation/SubmitChecklist4
-rw-r--r--Documentation/kdump/kdump.txt49
-rw-r--r--Documentation/pci.txt702
3 files changed, 591 insertions, 164 deletions
diff --git a/Documentation/SubmitChecklist b/Documentation/SubmitChecklist
index 2270efa10153..bfbb2718a279 100644
--- a/Documentation/SubmitChecklist
+++ b/Documentation/SubmitChecklist
@@ -72,3 +72,7 @@ kernel patches.
72 72
73 If the new code is substantial, addition of subsystem-specific fault 73 If the new code is substantial, addition of subsystem-specific fault
74 injection might be appropriate. 74 injection might be appropriate.
75
7622: Newly-added code has been compiled with `gcc -W'. This will generate
77 lots of noise, but is good for finding bugs like "warning: comparison
78 between signed and unsigned".
diff --git a/Documentation/kdump/kdump.txt b/Documentation/kdump/kdump.txt
index 5af6676a88f0..073306818347 100644
--- a/Documentation/kdump/kdump.txt
+++ b/Documentation/kdump/kdump.txt
@@ -17,7 +17,7 @@ You can use common Linux commands, such as cp and scp, to copy the
17memory image to a dump file on the local disk, or across the network to 17memory image to a dump file on the local disk, or across the network to
18a remote system. 18a remote system.
19 19
20Kdump and kexec are currently supported on the x86, x86_64, ppc64 and IA64 20Kdump and kexec are currently supported on the x86, x86_64, ppc64 and ia64
21architectures. 21architectures.
22 22
23When the system kernel boots, it reserves a small section of memory for 23When the system kernel boots, it reserves a small section of memory for
@@ -61,7 +61,12 @@ Install kexec-tools
61 61
622) Download the kexec-tools user-space package from the following URL: 622) Download the kexec-tools user-space package from the following URL:
63 63
64http://www.kernel.org/pub/linux/kernel/people/horms/kexec-tools/kexec-tools-testing-20061214.tar.gz 64http://www.kernel.org/pub/linux/kernel/people/horms/kexec-tools/kexec-tools-testing.tar.gz
65
66This is a symlink to the latest version, which at the time of writing is
6720061214, the only release of kexec-tools-testing so far. As other versions
68are made released, the older onese will remain available at
69http://www.kernel.org/pub/linux/kernel/people/horms/kexec-tools/
65 70
66Note: Latest kexec-tools-testing git tree is available at 71Note: Latest kexec-tools-testing git tree is available at
67 72
@@ -71,11 +76,11 @@ http://www.kernel.org/git/?p=linux/kernel/git/horms/kexec-tools-testing.git;a=su
71 76
723) Unpack the tarball with the tar command, as follows: 773) Unpack the tarball with the tar command, as follows:
73 78
74 tar xvpzf kexec-tools-testing-20061214.tar.gz 79 tar xvpzf kexec-tools-testing.tar.gz
75 80
764) Change to the kexec-tools-1.101 directory, as follows: 814) Change to the kexec-tools directory, as follows:
77 82
78 cd kexec-tools-testing-20061214 83 cd kexec-tools-testing-VERSION
79 84
805) Configure the package, as follows: 855) Configure the package, as follows:
81 86
@@ -224,7 +229,23 @@ Dump-capture kernel config options (Arch Dependent, ppc64)
224 229
225Dump-capture kernel config options (Arch Dependent, ia64) 230Dump-capture kernel config options (Arch Dependent, ia64)
226---------------------------------------------------------- 231----------------------------------------------------------
227(To be filled) 232
233- No specific options are required to create a dump-capture kernel
234 for ia64, other than those specified in the arch idependent section
235 above. This means that it is possible to use the system kernel
236 as a dump-capture kernel if desired.
237
238 The crashkernel region can be automatically placed by the system
239 kernel at run time. This is done by specifying the base address as 0,
240 or omitting it all together.
241
242 crashkernel=256M@0
243 or
244 crashkernel=256M
245
246 If the start address is specified, note that the start address of the
247 kernel will be aligned to 64Mb, so if the start address is not then
248 any space below the alignment point will be wasted.
228 249
229 250
230Boot into System Kernel 251Boot into System Kernel
@@ -243,6 +264,10 @@ Boot into System Kernel
243 264
244 On ppc64, use "crashkernel=128M@32M". 265 On ppc64, use "crashkernel=128M@32M".
245 266
267 On ia64, 256M@256M is a generous value that typically works.
268 The region may be automatically placed on ia64, see the
269 dump-capture kernel config option notes above.
270
246Load the Dump-capture Kernel 271Load the Dump-capture Kernel
247============================ 272============================
248 273
@@ -261,7 +286,8 @@ For x86_64:
261For ppc64: 286For ppc64:
262 - Use vmlinux 287 - Use vmlinux
263For ia64: 288For ia64:
264 (To be filled) 289 - Use vmlinux or vmlinuz.gz
290
265 291
266If you are using a uncompressed vmlinux image then use following command 292If you are using a uncompressed vmlinux image then use following command
267to load dump-capture kernel. 293to load dump-capture kernel.
@@ -277,18 +303,19 @@ to load dump-capture kernel.
277 --initrd=<initrd-for-dump-capture-kernel> \ 303 --initrd=<initrd-for-dump-capture-kernel> \
278 --append="root=<root-dev> <arch-specific-options>" 304 --append="root=<root-dev> <arch-specific-options>"
279 305
306Please note, that --args-linux does not need to be specified for ia64.
307It is planned to make this a no-op on that architecture, but for now
308it should be omitted
309
280Following are the arch specific command line options to be used while 310Following are the arch specific command line options to be used while
281loading dump-capture kernel. 311loading dump-capture kernel.
282 312
283For i386 and x86_64: 313For i386, x86_64 and ia64:
284 "init 1 irqpoll maxcpus=1" 314 "init 1 irqpoll maxcpus=1"
285 315
286For ppc64: 316For ppc64:
287 "init 1 maxcpus=1 noirqdistrib" 317 "init 1 maxcpus=1 noirqdistrib"
288 318
289For IA64
290 (To be filled)
291
292 319
293Notes on loading the dump-capture kernel: 320Notes on loading the dump-capture kernel:
294 321
diff --git a/Documentation/pci.txt b/Documentation/pci.txt
index 2b395e478961..fd5028eca13e 100644
--- a/Documentation/pci.txt
+++ b/Documentation/pci.txt
@@ -1,142 +1,231 @@
1 How To Write Linux PCI Drivers
2 1
3 by Martin Mares <mj@ucw.cz> on 07-Feb-2000 2 How To Write Linux PCI Drivers
3
4 by Martin Mares <mj@ucw.cz> on 07-Feb-2000
5 updated by Grant Grundler <grundler@parisc-linux.org> on 23-Dec-2006
4 6
5~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 7~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
6The world of PCI is vast and it's full of (mostly unpleasant) surprises. 8The world of PCI is vast and full of (mostly unpleasant) surprises.
7Different PCI devices have different requirements and different bugs -- 9Since each CPU architecture implements different chip-sets and PCI devices
8because of this, the PCI support layer in Linux kernel is not as trivial 10have different requirements (erm, "features"), the result is the PCI support
9as one would wish. This short pamphlet tries to help all potential driver 11in the Linux kernel is not as trivial as one would wish. This short paper
10authors find their way through the deep forests of PCI handling. 12tries to introduce all potential driver authors to Linux APIs for
13PCI device drivers.
14
15A more complete resource is the third edition of "Linux Device Drivers"
16by Jonathan Corbet, Alessandro Rubini, and Greg Kroah-Hartman.
17LDD3 is available for free (under Creative Commons License) from:
18
19 http://lwn.net/Kernel/LDD3/
20
21However, keep in mind that all documents are subject to "bit rot".
22Refer to the source code if things are not working as described here.
23
24Please send questions/comments/patches about Linux PCI API to the
25"Linux PCI" <linux-pci@atrey.karlin.mff.cuni.cz> mailing list.
26
11 27
12 28
130. Structure of PCI drivers 290. Structure of PCI drivers
14~~~~~~~~~~~~~~~~~~~~~~~~~~~ 30~~~~~~~~~~~~~~~~~~~~~~~~~~~
15There exist two kinds of PCI drivers: new-style ones (which leave most of 31PCI drivers "discover" PCI devices in a system via pci_register_driver().
16probing for devices to the PCI layer and support online insertion and removal 32Actually, it's the other way around. When the PCI generic code discovers
17of devices [thus supporting PCI, hot-pluggable PCI and CardBus in a single 33a new device, the driver with a matching "description" will be notified.
18driver]) and old-style ones which just do all the probing themselves. Unless 34Details on this below.
19you have a very good reason to do so, please don't use the old way of probing 35
20in any new code. After the driver finds the devices it wishes to operate 36pci_register_driver() leaves most of the probing for devices to
21on (either the old or the new way), it needs to perform the following steps: 37the PCI layer and supports online insertion/removal of devices [thus
38supporting hot-pluggable PCI, CardBus, and Express-Card in a single driver].
39pci_register_driver() call requires passing in a table of function
40pointers and thus dictates the high level structure of a driver.
41
42Once the driver knows about a PCI device and takes ownership, the
43driver generally needs to perform the following initialization:
22 44
23 Enable the device 45 Enable the device
24 Access device configuration space 46 Request MMIO/IOP resources
25 Discover resources (addresses and IRQ numbers) provided by the device 47 Set the DMA mask size (for both coherent and streaming DMA)
26 Allocate these resources 48 Allocate and initialize shared control data (pci_allocate_coherent())
27 Communicate with the device 49 Access device configuration space (if needed)
50 Register IRQ handler (request_irq())
51 Initialize non-PCI (i.e. LAN/SCSI/etc parts of the chip)
52 Enable DMA/processing engines
53
54When done using the device, and perhaps the module needs to be unloaded,
55the driver needs to take the follow steps:
56 Disable the device from generating IRQs
57 Release the IRQ (free_irq())
58 Stop all DMA activity
59 Release DMA buffers (both streaming and coherent)
60 Unregister from other subsystems (e.g. scsi or netdev)
61 Release MMIO/IOP resources
28 Disable the device 62 Disable the device
29 63
30Most of these topics are covered by the following sections, for the rest 64Most of these topics are covered in the following sections.
31look at <linux/pci.h>, it's hopefully well commented. 65For the rest look at LDD3 or <linux/pci.h> .
32 66
33If the PCI subsystem is not configured (CONFIG_PCI is not set), most of 67If the PCI subsystem is not configured (CONFIG_PCI is not set), most of
34the functions described below are defined as inline functions either completely 68the PCI functions described below are defined as inline functions either
35empty or just returning an appropriate error codes to avoid lots of ifdefs 69completely empty or just returning an appropriate error codes to avoid
36in the drivers. 70lots of ifdefs in the drivers.
71
37 72
38 73
391. New-style drivers 741. pci_register_driver() call
40~~~~~~~~~~~~~~~~~~~~ 75~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
41The new-style drivers just call pci_register_driver during their initialization
42with a pointer to a structure describing the driver (struct pci_driver) which
43contains:
44 76
45 name Name of the driver 77PCI device drivers call pci_register_driver() during their
78initialization with a pointer to a structure describing the driver
79(struct pci_driver):
80
81 field name Description
82 ---------- ------------------------------------------------------
46 id_table Pointer to table of device ID's the driver is 83 id_table Pointer to table of device ID's the driver is
47 interested in. Most drivers should export this 84 interested in. Most drivers should export this
48 table using MODULE_DEVICE_TABLE(pci,...). 85 table using MODULE_DEVICE_TABLE(pci,...).
49 probe Pointer to a probing function which gets called (during 86
50 execution of pci_register_driver for already existing 87 probe This probing function gets called (during execution
51 devices or later if a new device gets inserted) for all 88 of pci_register_driver() for already existing
52 PCI devices which match the ID table and are not handled 89 devices or later if a new device gets inserted) for
53 by the other drivers yet. This function gets passed a 90 all PCI devices which match the ID table and are not
54 pointer to the pci_dev structure representing the device 91 "owned" by the other drivers yet. This function gets
55 and also which entry in the ID table did the device 92 passed a "struct pci_dev *" for each device whose
56 match. It returns zero when the driver has accepted the 93 entry in the ID table matches the device. The probe
57 device or an error code (negative number) otherwise. 94 function returns zero when the driver chooses to
58 This function always gets called from process context, 95 take "ownership" of the device or an error code
59 so it can sleep. 96 (negative number) otherwise.
60 remove Pointer to a function which gets called whenever a 97 The probe function always gets called from process
61 device being handled by this driver is removed (either 98 context, so it can sleep.
62 during deregistration of the driver or when it's 99
63 manually pulled out of a hot-pluggable slot). This 100 remove The remove() function gets called whenever a device
64 function always gets called from process context, so it 101 being handled by this driver is removed (either during
65 can sleep. 102 deregistration of the driver or when it's manually
66 save_state Save a device's state before it's suspend. 103 pulled out of a hot-pluggable slot).
104 The remove function always gets called from process
105 context, so it can sleep.
106
67 suspend Put device into low power state. 107 suspend Put device into low power state.
108 suspend_late Put device into low power state.
109
110 resume_early Wake device from low power state.
68 resume Wake device from low power state. 111 resume Wake device from low power state.
112
113 (Please see Documentation/power/pci.txt for descriptions
114 of PCI Power Management and the related functions.)
115
69 enable_wake Enable device to generate wake events from a low power 116 enable_wake Enable device to generate wake events from a low power
70 state. 117 state.
71 118
72 (Please see Documentation/power/pci.txt for descriptions 119 shutdown Hook into reboot_notifier_list (kernel/sys.c).
73 of PCI Power Management and the related functions) 120 Intended to stop any idling DMA operations.
121 Useful for enabling wake-on-lan (NIC) or changing
122 the power state of a device before reboot.
123 e.g. drivers/net/e100.c.
124
125 err_handler See Documentation/pci-error-recovery.txt
126
127 multithread_probe Enable multi-threaded probe/scan. Driver must
128 provide its own locking/syncronization for init
129 operations if this is enabled.
130
74 131
75The ID table is an array of struct pci_device_id ending with a all-zero entry. 132The ID table is an array of struct pci_device_id entries ending with an
76Each entry consists of: 133all-zero entry. Each entry consists of:
134
135 vendor,device Vendor and device ID to match (or PCI_ANY_ID)
77 136
78 vendor, device Vendor and device ID to match (or PCI_ANY_ID)
79 subvendor, Subsystem vendor and device ID to match (or PCI_ANY_ID) 137 subvendor, Subsystem vendor and device ID to match (or PCI_ANY_ID)
80 subdevice 138 subdevice,
81 class, Device class to match. The class_mask tells which bits 139
82 class_mask of the class are honored during the comparison. 140 class Device class, subclass, and "interface" to match.
141 See Appendix D of the PCI Local Bus Spec or
142 include/linux/pci_ids.h for a full list of classes.
143 Most drivers do not need to specify class/class_mask
144 as vendor/device is normally sufficient.
145
146 class_mask limit which sub-fields of the class field are compared.
147 See drivers/scsi/sym53c8xx_2/ for example of usage.
148
83 driver_data Data private to the driver. 149 driver_data Data private to the driver.
150 Most drivers don't need to use driver_data field.
151 Best practice is to use driver_data as an index
152 into a static list of equivalent device types,
153 instead of using it as a pointer.
84 154
85Most drivers don't need to use the driver_data field. Best practice
86for use of driver_data is to use it as an index into a static list of
87equivalent device types, not to use it as a pointer.
88 155
89Have a table entry {PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID, PCI_ANY_ID} 156Most drivers only need PCI_DEVICE() or PCI_DEVICE_CLASS() to set up
90to have probe() called for every PCI device known to the system. 157a pci_device_id table.
91 158
92New PCI IDs may be added to a device driver at runtime by writing 159New PCI IDs may be added to a device driver pci_ids table at runtime
93to the file /sys/bus/pci/drivers/{driver}/new_id. When added, the 160as shown below:
94driver will probe for all devices it can support.
95 161
96echo "vendor device subvendor subdevice class class_mask driver_data" > \ 162echo "vendor device subvendor subdevice class class_mask driver_data" > \
97 /sys/bus/pci/drivers/{driver}/new_id 163/sys/bus/pci/drivers/{driver}/new_id
98where all fields are passed in as hexadecimal values (no leading 0x). 164
99Users need pass only as many fields as necessary; vendor, device, 165All fields are passed in as hexadecimal values (no leading 0x).
100subvendor, and subdevice fields default to PCI_ANY_ID (FFFFFFFF), 166Users need pass only as many fields as necessary:
101class and classmask fields default to 0, and driver_data defaults to 167 o vendor, device, subvendor, and subdevice fields default
1020UL. Device drivers must initialize use_driver_data in the dynids struct 168 to PCI_ANY_ID (FFFFFFFF),
103in their pci_driver struct prior to calling pci_register_driver in order 169 o class and classmask fields default to 0
104for the driver_data field to get passed to the driver. Otherwise, only a 170 o driver_data defaults to 0UL.
1050 is passed in that field. 171
172Once added, the driver probe routine will be invoked for any unclaimed
173PCI devices listed in its (newly updated) pci_ids list.
106 174
107When the driver exits, it just calls pci_unregister_driver() and the PCI layer 175When the driver exits, it just calls pci_unregister_driver() and the PCI layer
108automatically calls the remove hook for all devices handled by the driver. 176automatically calls the remove hook for all devices handled by the driver.
109 177
178
1791.1 "Attributes" for driver functions/data
180
110Please mark the initialization and cleanup functions where appropriate 181Please mark the initialization and cleanup functions where appropriate
111(the corresponding macros are defined in <linux/init.h>): 182(the corresponding macros are defined in <linux/init.h>):
112 183
113 __init Initialization code. Thrown away after the driver 184 __init Initialization code. Thrown away after the driver
114 initializes. 185 initializes.
115 __exit Exit code. Ignored for non-modular drivers. 186 __exit Exit code. Ignored for non-modular drivers.
116 __devinit Device initialization code. Identical to __init if 187
117 the kernel is not compiled with CONFIG_HOTPLUG, normal 188
118 function otherwise. 189 __devinit Device initialization code.
190 Identical to __init if the kernel is not compiled
191 with CONFIG_HOTPLUG, normal function otherwise.
119 __devexit The same for __exit. 192 __devexit The same for __exit.
120 193
121Tips: 194Tips on when/where to use the above attributes:
122 The module_init()/module_exit() functions (and all initialization 195 o The module_init()/module_exit() functions (and all
123 functions called only from these) should be marked __init/exit. 196 initialization functions called _only_ from these)
124 The struct pci_driver shouldn't be marked with any of these tags. 197 should be marked __init/__exit.
125 The ID table array should be marked __devinitdata.
126 The probe() and remove() functions (and all initialization
127 functions called only from these) should be marked __devinit/exit.
128 If you are sure the driver is not a hotplug driver then use only
129 __init/exit __initdata/exitdata.
130 198
131 Pointers to functions marked as __devexit must be created using 199 o Do not mark the struct pci_driver.
132 __devexit_p(function_name). That will generate the function
133 name or NULL if the __devexit function will be discarded.
134 200
201 o The ID table array should be marked __devinitdata.
135 202
1362. How to find PCI devices manually (the old style) 203 o The probe() and remove() functions should be marked __devinit
137~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 204 and __devexit respectively. All initialization functions
138PCI drivers not using the pci_register_driver() interface search 205 exclusively called by the probe() routine, can be marked __devinit.
139for PCI devices manually using the following constructs: 206 Ditto for remove() and __devexit.
207
208 o If mydriver_probe() is marked with __devinit(), then all address
209 references to mydriver_probe must use __devexit_p(mydriver_probe)
210 (in the struct pci_driver declaration for example).
211 __devexit_p() will generate the function name _or_ NULL if the
212 function will be discarded. For an example, see drivers/net/tg3.c.
213
214 o Do NOT mark a function if you are not sure which mark to use.
215 Better to not mark the function than mark the function wrong.
216
217
218
2192. How to find PCI devices manually
220~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
221
222PCI drivers should have a really good reason for not using the
223pci_register_driver() interface to search for PCI devices.
224The main reason PCI devices are controlled by multiple drivers
225is because one PCI device implements several different HW services.
226E.g. combined serial/parallel port/floppy controller.
227
228A manual search may be performed using the following constructs:
140 229
141Searching by vendor and device ID: 230Searching by vendor and device ID:
142 231
@@ -150,87 +239,311 @@ Searching by class ID (iterate in a similar way):
150 239
151Searching by both vendor/device and subsystem vendor/device ID: 240Searching by both vendor/device and subsystem vendor/device ID:
152 241
153 pci_get_subsys(VENDOR_ID, DEVICE_ID, SUBSYS_VENDOR_ID, SUBSYS_DEVICE_ID, dev). 242 pci_get_subsys(VENDOR_ID,DEVICE_ID, SUBSYS_VENDOR_ID, SUBSYS_DEVICE_ID, dev).
154 243
155 You can use the constant PCI_ANY_ID as a wildcard replacement for 244You can use the constant PCI_ANY_ID as a wildcard replacement for
156VENDOR_ID or DEVICE_ID. This allows searching for any device from a 245VENDOR_ID or DEVICE_ID. This allows searching for any device from a
157specific vendor, for example. 246specific vendor, for example.
158 247
159 These functions are hotplug-safe. They increment the reference count on 248These functions are hotplug-safe. They increment the reference count on
160the pci_dev that they return. You must eventually (possibly at module unload) 249the pci_dev that they return. You must eventually (possibly at module unload)
161decrement the reference count on these devices by calling pci_dev_put(). 250decrement the reference count on these devices by calling pci_dev_put().
162 251
163 252
1643. Enabling and disabling devices
165~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
166 Before you do anything with the device you've found, you need to enable
167it by calling pci_enable_device() which enables I/O and memory regions of
168the device, allocates an IRQ if necessary, assigns missing resources if
169needed and wakes up the device if it was in suspended state. Please note
170that this function can fail.
171 253
172 If you want to use the device in bus mastering mode, call pci_set_master() 2543. Device Initialization Steps
173which enables the bus master bit in PCI_COMMAND register and also fixes 255~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
174the latency timer value if it's set to something bogus by the BIOS. 256
257As noted in the introduction, most PCI drivers need the following steps
258for device initialization:
175 259
176 If you want to use the PCI Memory-Write-Invalidate transaction, 260 Enable the device
261 Request MMIO/IOP resources
262 Set the DMA mask size (for both coherent and streaming DMA)
263 Allocate and initialize shared control data (pci_allocate_coherent())
264 Access device configuration space (if needed)
265 Register IRQ handler (request_irq())
266 Initialize non-PCI (i.e. LAN/SCSI/etc parts of the chip)
267 Enable DMA/processing engines.
268
269The driver can access PCI config space registers at any time.
270(Well, almost. When running BIST, config space can go away...but
271that will just result in a PCI Bus Master Abort and config reads
272will return garbage).
273
274
2753.1 Enable the PCI device
276~~~~~~~~~~~~~~~~~~~~~~~~~
277Before touching any device registers, the driver needs to enable
278the PCI device by calling pci_enable_device(). This will:
279 o wake up the device if it was in suspended state,
280 o allocate I/O and memory regions of the device (if BIOS did not),
281 o allocate an IRQ (if BIOS did not).
282
283NOTE: pci_enable_device() can fail! Check the return value.
284NOTE2: Also see pci_enable_device_bars() below. Drivers can
285 attempt to enable only a subset of BARs they need.
286
287[ OS BUG: we don't check resource allocations before enabling those
288 resources. The sequence would make more sense if we called
289 pci_request_resources() before calling pci_enable_device().
290 Currently, the device drivers can't detect the bug when when two
291 devices have been allocated the same range. This is not a common
292 problem and unlikely to get fixed soon.
293
294 This has been discussed before but not changed as of 2.6.19:
295 http://lkml.org/lkml/2006/3/2/194
296]
297
298pci_set_master() will enable DMA by setting the bus master bit
299in the PCI_COMMAND register. It also fixes the latency timer value if
300it's set to something bogus by the BIOS.
301
302If the PCI device can use the PCI Memory-Write-Invalidate transaction,
177call pci_set_mwi(). This enables the PCI_COMMAND bit for Mem-Wr-Inval 303call pci_set_mwi(). This enables the PCI_COMMAND bit for Mem-Wr-Inval
178and also ensures that the cache line size register is set correctly. 304and also ensures that the cache line size register is set correctly.
179Make sure to check the return value of pci_set_mwi(), not all architectures 305Check the return value of pci_set_mwi() as not all architectures
180may support Memory-Write-Invalidate. 306or chip-sets may support Memory-Write-Invalidate.
307
308
3093.2 Request MMIO/IOP resources
310~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
311Memory (MMIO), and I/O port addresses should NOT be read directly
312from the PCI device config space. Use the values in the pci_dev structure
313as the PCI "bus address" might have been remapped to a "host physical"
314address by the arch/chip-set specific kernel support.
181 315
182 If your driver decides to stop using the device (e.g., there was an 316See Documentation/IO-mapping.txt for how to access device registers
183error while setting it up or the driver module is being unloaded), it 317or device memory.
184should call pci_disable_device() to deallocate any IRQ resources, disable 318
185PCI bus-mastering, etc. You should not do anything with the device after 319The device driver needs to call pci_request_region() to verify
320no other device is already using the same address resource.
321Conversely, drivers should call pci_release_region() AFTER
186calling pci_disable_device(). 322calling pci_disable_device().
323The idea is to prevent two devices colliding on the same address range.
324
325[ See OS BUG comment above. Currently (2.6.19), The driver can only
326 determine MMIO and IO Port resource availability _after_ calling
327 pci_enable_device(). ]
328
329Generic flavors of pci_request_region() are request_mem_region()
330(for MMIO ranges) and request_region() (for IO Port ranges).
331Use these for address resources that are not described by "normal" PCI
332BARs.
333
334Also see pci_request_selected_regions() below.
335
336
3373.3 Set the DMA mask size
338~~~~~~~~~~~~~~~~~~~~~~~~~
339[ If anything below doesn't make sense, please refer to
340 Documentation/DMA-API.txt. This section is just a reminder that
341 drivers need to indicate DMA capabilities of the device and is not
342 an authoritative source for DMA interfaces. ]
343
344While all drivers should explicitly indicate the DMA capability
345(e.g. 32 or 64 bit) of the PCI bus master, devices with more than
34632-bit bus master capability for streaming data need the driver
347to "register" this capability by calling pci_set_dma_mask() with
348appropriate parameters. In general this allows more efficient DMA
349on systems where System RAM exists above 4G _physical_ address.
350
351Drivers for all PCI-X and PCIe compliant devices must call
352pci_set_dma_mask() as they are 64-bit DMA devices.
353
354Similarly, drivers must also "register" this capability if the device
355can directly address "consistent memory" in System RAM above 4G physical
356address by calling pci_set_consistent_dma_mask().
357Again, this includes drivers for all PCI-X and PCIe compliant devices.
358Many 64-bit "PCI" devices (before PCI-X) and some PCI-X devices are
35964-bit DMA capable for payload ("streaming") data but not control
360("consistent") data.
361
362
3633.4 Setup shared control data
364~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
365Once the DMA masks are set, the driver can allocate "consistent" (a.k.a. shared)
366memory. See Documentation/DMA-API.txt for a full description of
367the DMA APIs. This section is just a reminder that it needs to be done
368before enabling DMA on the device.
369
370
3713.5 Initialize device registers
372~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
373Some drivers will need specific "capability" fields programmed
374or other "vendor specific" register initialized or reset.
375E.g. clearing pending interrupts.
376
377
3783.6 Register IRQ handler
379~~~~~~~~~~~~~~~~~~~~~~~~
380While calling request_irq() is the the last step described here,
381this is often just another intermediate step to initialize a device.
382This step can often be deferred until the device is opened for use.
383
384All interrupt handlers for IRQ lines should be registered with IRQF_SHARED
385and use the devid to map IRQs to devices (remember that all PCI IRQ lines
386can be shared).
387
388request_irq() will associate an interrupt handler and device handle
389with an interrupt number. Historically interrupt numbers represent
390IRQ lines which run from the PCI device to the Interrupt controller.
391With MSI and MSI-X (more below) the interrupt number is a CPU "vector".
392
393request_irq() also enables the interrupt. Make sure the device is
394quiesced and does not have any interrupts pending before registering
395the interrupt handler.
396
397MSI and MSI-X are PCI capabilities. Both are "Message Signaled Interrupts"
398which deliver interrupts to the CPU via a DMA write to a Local APIC.
399The fundamental difference between MSI and MSI-X is how multiple
400"vectors" get allocated. MSI requires contiguous blocks of vectors
401while MSI-X can allocate several individual ones.
402
403MSI capability can be enabled by calling pci_enable_msi() or
404pci_enable_msix() before calling request_irq(). This causes
405the PCI support to program CPU vector data into the PCI device
406capability registers.
407
408If your PCI device supports both, try to enable MSI-X first.
409Only one can be enabled at a time. Many architectures, chip-sets,
410or BIOSes do NOT support MSI or MSI-X and the call to pci_enable_msi/msix
411will fail. This is important to note since many drivers have
412two (or more) interrupt handlers: one for MSI/MSI-X and another for IRQs.
413They choose which handler to register with request_irq() based on the
414return value from pci_enable_msi/msix().
415
416There are (at least) two really good reasons for using MSI:
4171) MSI is an exclusive interrupt vector by definition.
418 This means the interrupt handler doesn't have to verify
419 its device caused the interrupt.
420
4212) MSI avoids DMA/IRQ race conditions. DMA to host memory is guaranteed
422 to be visible to the host CPU(s) when the MSI is delivered. This
423 is important for both data coherency and avoiding stale control data.
424 This guarantee allows the driver to omit MMIO reads to flush
425 the DMA stream.
426
427See drivers/infiniband/hw/mthca/ or drivers/net/tg3.c for examples
428of MSI/MSI-X usage.
429
430
431
4324. PCI device shutdown
433~~~~~~~~~~~~~~~~~~~~~~~
434
435When a PCI device driver is being unloaded, most of the following
436steps need to be performed:
437
438 Disable the device from generating IRQs
439 Release the IRQ (free_irq())
440 Stop all DMA activity
441 Release DMA buffers (both streaming and consistent)
442 Unregister from other subsystems (e.g. scsi or netdev)
443 Disable device from responding to MMIO/IO Port addresses
444 Release MMIO/IO Port resource(s)
445
446
4474.1 Stop IRQs on the device
448~~~~~~~~~~~~~~~~~~~~~~~~~~~
449How to do this is chip/device specific. If it's not done, it opens
450the possibility of a "screaming interrupt" if (and only if)
451the IRQ is shared with another device.
452
453When the shared IRQ handler is "unhooked", the remaining devices
454using the same IRQ line will still need the IRQ enabled. Thus if the
455"unhooked" device asserts IRQ line, the system will respond assuming
456it was one of the remaining devices asserted the IRQ line. Since none
457of the other devices will handle the IRQ, the system will "hang" until
458it decides the IRQ isn't going to get handled and masks the IRQ (100,000
459iterations later). Once the shared IRQ is masked, the remaining devices
460will stop functioning properly. Not a nice situation.
461
462This is another reason to use MSI or MSI-X if it's available.
463MSI and MSI-X are defined to be exclusive interrupts and thus
464are not susceptible to the "screaming interrupt" problem.
465
466
4674.2 Release the IRQ
468~~~~~~~~~~~~~~~~~~~
469Once the device is quiesced (no more IRQs), one can call free_irq().
470This function will return control once any pending IRQs are handled,
471"unhook" the drivers IRQ handler from that IRQ, and finally release
472the IRQ if no one else is using it.
473
474
4754.3 Stop all DMA activity
476~~~~~~~~~~~~~~~~~~~~~~~~~
477It's extremely important to stop all DMA operations BEFORE attempting
478to deallocate DMA control data. Failure to do so can result in memory
479corruption, hangs, and on some chip-sets a hard crash.
187 480
1884. How to access PCI config space 481Stopping DMA after stopping the IRQs can avoid races where the
482IRQ handler might restart DMA engines.
483
484While this step sounds obvious and trivial, several "mature" drivers
485didn't get this step right in the past.
486
487
4884.4 Release DMA buffers
489~~~~~~~~~~~~~~~~~~~~~~~
490Once DMA is stopped, clean up streaming DMA first.
491I.e. unmap data buffers and return buffers to "upstream"
492owners if there is one.
493
494Then clean up "consistent" buffers which contain the control data.
495
496See Documentation/DMA-API.txt for details on unmapping interfaces.
497
498
4994.5 Unregister from other subsystems
500~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
501Most low level PCI device drivers support some other subsystem
502like USB, ALSA, SCSI, NetDev, Infiniband, etc. Make sure your
503driver isn't losing resources from that other subsystem.
504If this happens, typically the symptom is an Oops (panic) when
505the subsystem attempts to call into a driver that has been unloaded.
506
507
5084.6 Disable Device from responding to MMIO/IO Port addresses
509~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
510io_unmap() MMIO or IO Port resources and then call pci_disable_device().
511This is the symmetric opposite of pci_enable_device().
512Do not access device registers after calling pci_disable_device().
513
514
5154.7 Release MMIO/IO Port Resource(s)
516~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
517Call pci_release_region() to mark the MMIO or IO Port range as available.
518Failure to do so usually results in the inability to reload the driver.
519
520
521
5225. How to access PCI config space
189~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 523~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
190 You can use pci_(read|write)_config_(byte|word|dword) to access the config 524
525You can use pci_(read|write)_config_(byte|word|dword) to access the config
191space of a device represented by struct pci_dev *. All these functions return 0 526space of a device represented by struct pci_dev *. All these functions return 0
192when successful or an error code (PCIBIOS_...) which can be translated to a text 527when successful or an error code (PCIBIOS_...) which can be translated to a text
193string by pcibios_strerror. Most drivers expect that accesses to valid PCI 528string by pcibios_strerror. Most drivers expect that accesses to valid PCI
194devices don't fail. 529devices don't fail.
195 530
196 If you don't have a struct pci_dev available, you can call 531If you don't have a struct pci_dev available, you can call
197pci_bus_(read|write)_config_(byte|word|dword) to access a given device 532pci_bus_(read|write)_config_(byte|word|dword) to access a given device
198and function on that bus. 533and function on that bus.
199 534
200 If you access fields in the standard portion of the config header, please 535If you access fields in the standard portion of the config header, please
201use symbolic names of locations and bits declared in <linux/pci.h>. 536use symbolic names of locations and bits declared in <linux/pci.h>.
202 537
203 If you need to access Extended PCI Capability registers, just call 538If you need to access Extended PCI Capability registers, just call
204pci_find_capability() for the particular capability and it will find the 539pci_find_capability() for the particular capability and it will find the
205corresponding register block for you. 540corresponding register block for you.
206 541
207 542
2085. Addresses and interrupts
209~~~~~~~~~~~~~~~~~~~~~~~~~~~
210 Memory and port addresses and interrupt numbers should NOT be read from the
211config space. You should use the values in the pci_dev structure as they might
212have been remapped by the kernel.
213
214 See Documentation/IO-mapping.txt for how to access device memory.
215
216 The device driver needs to call pci_request_region() to make sure
217no other device is already using the same resource. The driver is expected
218to determine MMIO and IO Port resource availability _before_ calling
219pci_enable_device(). Conversely, drivers should call pci_release_region()
220_after_ calling pci_disable_device(). The idea is to prevent two devices
221colliding on the same address range.
222
223Generic flavors of pci_request_region() are request_mem_region()
224(for MMIO ranges) and request_region() (for IO Port ranges).
225Use these for address resources that are not described by "normal" PCI
226interfaces (e.g. BAR).
227
228 All interrupt handlers should be registered with IRQF_SHARED and use the devid
229to map IRQs to devices (remember that all PCI interrupts are shared).
230
231 543
2326. Other interesting functions 5446. Other interesting functions
233~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 545~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
546
234pci_find_slot() Find pci_dev corresponding to given bus and 547pci_find_slot() Find pci_dev corresponding to given bus and
235 slot numbers. 548 slot numbers.
236pci_set_power_state() Set PCI Power Management state (0=D0 ... 3=D3) 549pci_set_power_state() Set PCI Power Management state (0=D0 ... 3=D3)
@@ -247,11 +560,12 @@ pci_set_mwi() Enable Memory-Write-Invalidate transactions.
247pci_clear_mwi() Disable Memory-Write-Invalidate transactions. 560pci_clear_mwi() Disable Memory-Write-Invalidate transactions.
248 561
249 562
563
2507. Miscellaneous hints 5647. Miscellaneous hints
251~~~~~~~~~~~~~~~~~~~~~~ 565~~~~~~~~~~~~~~~~~~~~~~
252When displaying PCI slot names to the user (for example when a driver wants 566
253to tell the user what card has it found), please use pci_name(pci_dev) 567When displaying PCI device names to the user (for example when a driver wants
254for this purpose. 568to tell the user what card has it found), please use pci_name(pci_dev).
255 569
256Always refer to the PCI devices by a pointer to the pci_dev structure. 570Always refer to the PCI devices by a pointer to the pci_dev structure.
257All PCI layer functions use this identification and it's the only 571All PCI layer functions use this identification and it's the only
@@ -259,31 +573,113 @@ reasonable one. Don't use bus/slot/function numbers except for very
259special purposes -- on systems with multiple primary buses their semantics 573special purposes -- on systems with multiple primary buses their semantics
260can be pretty complex. 574can be pretty complex.
261 575
262If you're going to use PCI bus mastering DMA, take a look at
263Documentation/DMA-mapping.txt.
264
265Don't try to turn on Fast Back to Back writes in your driver. All devices 576Don't try to turn on Fast Back to Back writes in your driver. All devices
266on the bus need to be capable of doing it, so this is something which needs 577on the bus need to be capable of doing it, so this is something which needs
267to be handled by platform and generic code, not individual drivers. 578to be handled by platform and generic code, not individual drivers.
268 579
269 580
581
2708. Vendor and device identifications 5828. Vendor and device identifications
271~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 583~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
272For the future, let's avoid adding device ids to include/linux/pci_ids.h.
273 584
274PCI_VENDOR_ID_xxx for vendors, and a hex constant for device ids. 585One is not not required to add new device ids to include/linux/pci_ids.h.
586Please add PCI_VENDOR_ID_xxx for vendors and a hex constant for device ids.
587
588PCI_VENDOR_ID_xxx constants are re-used. The device ids are arbitrary
589hex numbers (vendor controlled) and normally used only in a single
590location, the pci_device_id table.
591
592Please DO submit new vendor/device ids to pciids.sourceforge.net project.
593
275 594
276Rationale: PCI_VENDOR_ID_xxx constants are re-used, but device ids are not.
277 Further, device ids are arbitrary hex numbers, normally used only in a
278 single location, the pci_device_id table.
279 595
2809. Obsolete functions 5969. Obsolete functions
281~~~~~~~~~~~~~~~~~~~~~ 597~~~~~~~~~~~~~~~~~~~~~
598
282There are several functions which you might come across when trying to 599There are several functions which you might come across when trying to
283port an old driver to the new PCI interface. They are no longer present 600port an old driver to the new PCI interface. They are no longer present
284in the kernel as they aren't compatible with hotplug or PCI domains or 601in the kernel as they aren't compatible with hotplug or PCI domains or
285having sane locking. 602having sane locking.
286 603
287pci_find_device() Superseded by pci_get_device() 604pci_find_device() Superseded by pci_get_device()
288pci_find_subsys() Superseded by pci_get_subsys() 605pci_find_subsys() Superseded by pci_get_subsys()
289pci_find_slot() Superseded by pci_get_slot() 606pci_find_slot() Superseded by pci_get_slot()
607
608
609The alternative is the traditional PCI device driver that walks PCI
610device lists. This is still possible but discouraged.
611
612
613
61410. pci_enable_device_bars() and Legacy I/O Port space
615~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
616
617Large servers may not be able to provide I/O port resources to all PCI
618devices. I/O Port space is only 64KB on Intel Architecture[1] and is
619likely also fragmented since the I/O base register of PCI-to-PCI
620bridge will usually be aligned to a 4KB boundary[2]. On such systems,
621pci_enable_device() and pci_request_region() will fail when
622attempting to enable I/O Port regions that don't have I/O Port
623resources assigned.
624
625Fortunately, many PCI devices which request I/O Port resources also
626provide access to the same registers via MMIO BARs. These devices can
627be handled without using I/O port space and the drivers typically
628offer a CONFIG_ option to only use MMIO regions
629(e.g. CONFIG_TULIP_MMIO). PCI devices typically provide I/O port
630interface for legacy OSes and will work when I/O port resources are not
631assigned. The "PCI Local Bus Specification Revision 3.0" discusses
632this on p.44, "IMPLEMENTATION NOTE".
633
634If your PCI device driver doesn't need I/O port resources assigned to
635I/O Port BARs, you should use pci_enable_device_bars() instead of
636pci_enable_device() in order not to enable I/O port regions for the
637corresponding devices. In addition, you should use
638pci_request_selected_regions() and pci_release_selected_regions()
639instead of pci_request_regions()/pci_release_regions() in order not to
640request/release I/O port regions for the corresponding devices.
641
642[1] Some systems support 64KB I/O port space per PCI segment.
643[2] Some PCI-to-PCI bridges support optional 1KB aligned I/O base.
644
645
646
64711. MMIO Space and "Write Posting"
648~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
649
650Converting a driver from using I/O Port space to using MMIO space
651often requires some additional changes. Specifically, "write posting"
652needs to be handled. Many drivers (e.g. tg3, acenic, sym53c8xx_2)
653already do this. I/O Port space guarantees write transactions reach the PCI
654device before the CPU can continue. Writes to MMIO space allow the CPU
655to continue before the transaction reaches the PCI device. HW weenies
656call this "Write Posting" because the write completion is "posted" to
657the CPU before the transaction has reached its destination.
658
659Thus, timing sensitive code should add readl() where the CPU is
660expected to wait before doing other work. The classic "bit banging"
661sequence works fine for I/O Port space:
662
663 for (i = 8; --i; val >>= 1) {
664 outb(val & 1, ioport_reg); /* write bit */
665 udelay(10);
666 }
667
668The same sequence for MMIO space should be:
669
670 for (i = 8; --i; val >>= 1) {
671 writeb(val & 1, mmio_reg); /* write bit */
672 readb(safe_mmio_reg); /* flush posted write */
673 udelay(10);
674 }
675
676It is important that "safe_mmio_reg" not have any side effects that
677interferes with the correct operation of the device.
678
679Another case to watch out for is when resetting a PCI device. Use PCI
680Configuration space reads to flush the writel(). This will gracefully
681handle the PCI master abort on all platforms if the PCI device is
682expected to not respond to a readl(). Most x86 platforms will allow
683MMIO reads to master abort (a.k.a. "Soft Fail") and return garbage
684(e.g. ~0). But many RISC platforms will crash (a.k.a."Hard Fail").
685