1 .. SPDX-License-Identifier: GPL-2.0 1 .. SPDX-License-Identifier: GPL-2.0
2 .. include:: <isonum.txt> 2 .. include:: <isonum.txt>
3 3
4 ============================================== 4 ===========================================================
5 The PCI Express Advanced Error Reporting Drive 5 The PCI Express Advanced Error Reporting Driver Guide HOWTO
6 ============================================== 6 ===========================================================
7 7
8 :Authors: - T. Long Nguyen <tom.l.nguyen@intel. 8 :Authors: - T. Long Nguyen <tom.l.nguyen@intel.com>
9 - Yanmin Zhang <yanmin.zhang@intel.co 9 - Yanmin Zhang <yanmin.zhang@intel.com>
10 10
11 :Copyright: |copy| 2006 Intel Corporation 11 :Copyright: |copy| 2006 Intel Corporation
12 12
13 Overview 13 Overview
14 =========== 14 ===========
15 15
16 About this guide 16 About this guide
17 ---------------- 17 ----------------
18 18
19 This guide describes the basics of the PCI Exp 19 This guide describes the basics of the PCI Express (PCIe) Advanced Error
20 Reporting (AER) driver and provides informatio 20 Reporting (AER) driver and provides information on how to use it, as
21 well as how to enable the drivers of Endpoint 21 well as how to enable the drivers of Endpoint devices to conform with
22 the PCIe AER driver. 22 the PCIe AER driver.
23 23
24 24
25 What is the PCIe AER Driver? 25 What is the PCIe AER Driver?
26 ---------------------------- 26 ----------------------------
27 27
28 PCIe error signaling can occur on the PCIe lin 28 PCIe error signaling can occur on the PCIe link itself
29 or on behalf of transactions initiated on the 29 or on behalf of transactions initiated on the link. PCIe
30 defines two error reporting paradigms: the bas 30 defines two error reporting paradigms: the baseline capability and
31 the Advanced Error Reporting capability. The b 31 the Advanced Error Reporting capability. The baseline capability is
32 required of all PCIe components providing a mi 32 required of all PCIe components providing a minimum defined
33 set of error reporting requirements. Advanced 33 set of error reporting requirements. Advanced Error Reporting
34 capability is implemented with a PCIe Advanced 34 capability is implemented with a PCIe Advanced Error Reporting
35 extended capability structure providing more r 35 extended capability structure providing more robust error reporting.
36 36
37 The PCIe AER driver provides the infrastructur 37 The PCIe AER driver provides the infrastructure to support PCIe Advanced
38 Error Reporting capability. The PCIe AER drive 38 Error Reporting capability. The PCIe AER driver provides three basic
39 functions: 39 functions:
40 40
41 - Gathers the comprehensive error informatio 41 - Gathers the comprehensive error information if errors occurred.
42 - Reports error to the users. 42 - Reports error to the users.
43 - Performs error recovery actions. 43 - Performs error recovery actions.
44 44
45 The AER driver only attaches to Root Ports and 45 The AER driver only attaches to Root Ports and RCECs that support the PCIe
46 AER capability. 46 AER capability.
47 47
48 48
49 User Guide 49 User Guide
50 ========== 50 ==========
51 51
52 Include the PCIe AER Root Driver into the Linu 52 Include the PCIe AER Root Driver into the Linux Kernel
53 ---------------------------------------------- 53 ------------------------------------------------------
54 54
55 The PCIe AER driver is a Root Port service dri 55 The PCIe AER driver is a Root Port service driver attached
56 via the PCIe Port Bus driver. If a user wants 56 via the PCIe Port Bus driver. If a user wants to use it, the driver
57 must be compiled. It is enabled with CONFIG_PC 57 must be compiled. It is enabled with CONFIG_PCIEAER, which
58 depends on CONFIG_PCIEPORTBUS. 58 depends on CONFIG_PCIEPORTBUS.
59 59
60 Load PCIe AER Root Driver 60 Load PCIe AER Root Driver
61 ------------------------- 61 -------------------------
62 62
63 Some systems have AER support in firmware. Ena 63 Some systems have AER support in firmware. Enabling Linux AER support at
64 the same time the firmware handles AER would r 64 the same time the firmware handles AER would result in unpredictable
65 behavior. Therefore, Linux does not handle AER 65 behavior. Therefore, Linux does not handle AER events unless the firmware
66 grants AER control to the OS via the ACPI _OSC 66 grants AER control to the OS via the ACPI _OSC method. See the PCI Firmware
67 Specification for details regarding _OSC usage 67 Specification for details regarding _OSC usage.
68 68
69 AER error output 69 AER error output
70 ---------------- 70 ----------------
71 71
72 When a PCIe AER error is captured, an error me 72 When a PCIe AER error is captured, an error message will be output to
73 console. If it's a correctable error, it is ou 73 console. If it's a correctable error, it is output as an info message.
74 Otherwise, it is printed as an error. So users 74 Otherwise, it is printed as an error. So users could choose different
75 log level to filter out correctable error mess 75 log level to filter out correctable error messages.
76 76
77 Below shows an example:: 77 Below shows an example::
78 78
79 0000:50:00.0: PCIe Bus Error: severity=Uncor 79 0000:50:00.0: PCIe Bus Error: severity=Uncorrected (Fatal), type=Transaction Layer, id=0500(Requester ID)
80 0000:50:00.0: device [8086:0329] error sta 80 0000:50:00.0: device [8086:0329] error status/mask=00100000/00000000
81 0000:50:00.0: [20] Unsupported Request 81 0000:50:00.0: [20] Unsupported Request (First)
82 0000:50:00.0: TLP Header: 04000001 00200a0 82 0000:50:00.0: TLP Header: 04000001 00200a03 05010000 00050100
83 83
84 In the example, 'Requester ID' means the ID of 84 In the example, 'Requester ID' means the ID of the device that sent
85 the error message to the Root Port. Please ref 85 the error message to the Root Port. Please refer to PCIe specs for other
86 fields. 86 fields.
87 87
88 AER Statistics / Counters 88 AER Statistics / Counters
89 ------------------------- 89 -------------------------
90 90
91 When PCIe AER errors are captured, the counter 91 When PCIe AER errors are captured, the counters / statistics are also exposed
92 in the form of sysfs attributes which are docu 92 in the form of sysfs attributes which are documented at
93 Documentation/ABI/testing/sysfs-bus-pci-device 93 Documentation/ABI/testing/sysfs-bus-pci-devices-aer_stats
94 94
95 Developer Guide 95 Developer Guide
96 =============== 96 ===============
97 97
98 To enable error recovery, a software driver mu 98 To enable error recovery, a software driver must provide callbacks.
99 99
100 To support AER better, developers need to unde 100 To support AER better, developers need to understand how AER works.
101 101
102 PCIe errors are classified into two types: cor 102 PCIe errors are classified into two types: correctable errors
103 and uncorrectable errors. This classification 103 and uncorrectable errors. This classification is based on the impact
104 of those errors, which may result in degraded 104 of those errors, which may result in degraded performance or function
105 failure. 105 failure.
106 106
107 Correctable errors pose no impacts on the func 107 Correctable errors pose no impacts on the functionality of the
108 interface. The PCIe protocol can recover witho 108 interface. The PCIe protocol can recover without any software
109 intervention or any loss of data. These errors 109 intervention or any loss of data. These errors are detected and
110 corrected by hardware. 110 corrected by hardware.
111 111
112 Unlike correctable errors, uncorrectable 112 Unlike correctable errors, uncorrectable
113 errors impact functionality of the interface. 113 errors impact functionality of the interface. Uncorrectable errors
114 can cause a particular transaction or a partic 114 can cause a particular transaction or a particular PCIe link
115 to be unreliable. Depending on those error con 115 to be unreliable. Depending on those error conditions, uncorrectable
116 errors are further classified into non-fatal e 116 errors are further classified into non-fatal errors and fatal errors.
117 Non-fatal errors cause the particular transact 117 Non-fatal errors cause the particular transaction to be unreliable,
118 but the PCIe link itself is fully functional. 118 but the PCIe link itself is fully functional. Fatal errors, on
119 the other hand, cause the link to be unreliabl 119 the other hand, cause the link to be unreliable.
120 120
121 When PCIe error reporting is enabled, a device 121 When PCIe error reporting is enabled, a device will automatically send an
122 error message to the Root Port above it when i 122 error message to the Root Port above it when it captures
123 an error. The Root Port, upon receiving an err 123 an error. The Root Port, upon receiving an error reporting message,
124 internally processes and logs the error messag 124 internally processes and logs the error message in its AER
125 Capability structure. Error information being 125 Capability structure. Error information being logged includes storing
126 the error reporting agent's requestor ID into 126 the error reporting agent's requestor ID into the Error Source
127 Identification Registers and setting the error 127 Identification Registers and setting the error bits of the Root Error
128 Status Register accordingly. If AER error repo 128 Status Register accordingly. If AER error reporting is enabled in the Root
129 Error Command Register, the Root Port generate 129 Error Command Register, the Root Port generates an interrupt when an
130 error is detected. 130 error is detected.
131 131
132 Note that the errors as described above are re 132 Note that the errors as described above are related to the PCIe
133 hierarchy and links. These errors do not inclu 133 hierarchy and links. These errors do not include any device specific
134 errors because device specific errors will sti 134 errors because device specific errors will still get sent directly to
135 the device driver. 135 the device driver.
136 136
137 Provide callbacks 137 Provide callbacks
138 ----------------- 138 -----------------
139 139
140 callback reset_link to reset PCIe link 140 callback reset_link to reset PCIe link
141 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 141 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
142 142
143 This callback is used to reset the PCIe physic 143 This callback is used to reset the PCIe physical link when a
144 fatal error happens. The Root Port AER service 144 fatal error happens. The Root Port AER service driver provides a
145 default reset_link function, but different Ups 145 default reset_link function, but different Upstream Ports might
146 have different specifications to reset the PCI 146 have different specifications to reset the PCIe link, so
147 Upstream Port drivers may provide their own re 147 Upstream Port drivers may provide their own reset_link functions.
148 148
149 Section 3.2.2.2 provides more detailed info on 149 Section 3.2.2.2 provides more detailed info on when to call
150 reset_link. 150 reset_link.
151 151
152 PCI error-recovery callbacks 152 PCI error-recovery callbacks
153 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 153 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
154 154
155 The PCIe AER Root driver uses error callbacks 155 The PCIe AER Root driver uses error callbacks to coordinate
156 with downstream device drivers associated with 156 with downstream device drivers associated with a hierarchy in question
157 when performing error recovery actions. 157 when performing error recovery actions.
158 158
159 Data struct pci_driver has a pointer, err_hand 159 Data struct pci_driver has a pointer, err_handler, to point to
160 pci_error_handlers who consists of a couple of 160 pci_error_handlers who consists of a couple of callback function
161 pointers. The AER driver follows the rules def 161 pointers. The AER driver follows the rules defined in
162 pci-error-recovery.rst except PCIe-specific pa 162 pci-error-recovery.rst except PCIe-specific parts (e.g.
163 reset_link). Please refer to pci-error-recover 163 reset_link). Please refer to pci-error-recovery.rst for detailed
164 definitions of the callbacks. 164 definitions of the callbacks.
165 165
166 The sections below specify when to call the er 166 The sections below specify when to call the error callback functions.
167 167
168 Correctable errors 168 Correctable errors
169 ~~~~~~~~~~~~~~~~~~ 169 ~~~~~~~~~~~~~~~~~~
170 170
171 Correctable errors pose no impacts on the func 171 Correctable errors pose no impacts on the functionality of
172 the interface. The PCIe protocol can recover w 172 the interface. The PCIe protocol can recover without any
173 software intervention or any loss of data. The 173 software intervention or any loss of data. These errors do not
174 require any recovery actions. The AER driver c 174 require any recovery actions. The AER driver clears the device's
175 correctable error status register accordingly 175 correctable error status register accordingly and logs these errors.
176 176
177 Non-correctable (non-fatal and fatal) errors 177 Non-correctable (non-fatal and fatal) errors
178 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 178 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
179 179
180 If an error message indicates a non-fatal erro 180 If an error message indicates a non-fatal error, performing link reset
181 at upstream is not required. The AER driver ca 181 at upstream is not required. The AER driver calls error_detected(dev,
182 pci_channel_io_normal) to all drivers associat 182 pci_channel_io_normal) to all drivers associated within a hierarchy in
183 question. For example:: 183 question. For example::
184 184
185 Endpoint <==> Downstream Port B <==> Upstrea 185 Endpoint <==> Downstream Port B <==> Upstream Port A <==> Root Port
186 186
187 If Upstream Port A captures an AER error, the 187 If Upstream Port A captures an AER error, the hierarchy consists of
188 Downstream Port B and Endpoint. 188 Downstream Port B and Endpoint.
189 189
190 A driver may return PCI_ERS_RESULT_CAN_RECOVER 190 A driver may return PCI_ERS_RESULT_CAN_RECOVER,
191 PCI_ERS_RESULT_DISCONNECT, or PCI_ERS_RESULT_N 191 PCI_ERS_RESULT_DISCONNECT, or PCI_ERS_RESULT_NEED_RESET, depending on
192 whether it can recover or the AER driver calls 192 whether it can recover or the AER driver calls mmio_enabled as next.
193 193
194 If an error message indicates a fatal error, k 194 If an error message indicates a fatal error, kernel will broadcast
195 error_detected(dev, pci_channel_io_frozen) to 195 error_detected(dev, pci_channel_io_frozen) to all drivers within
196 a hierarchy in question. Then, performing link 196 a hierarchy in question. Then, performing link reset at upstream is
197 necessary. As different kinds of devices might 197 necessary. As different kinds of devices might use different approaches
198 to reset link, AER port service driver is requ 198 to reset link, AER port service driver is required to provide the
199 function to reset link via callback parameter 199 function to reset link via callback parameter of pcie_do_recovery()
200 function. If reset_link is not NULL, recovery 200 function. If reset_link is not NULL, recovery function will use it
201 to reset the link. If error_detected returns P 201 to reset the link. If error_detected returns PCI_ERS_RESULT_CAN_RECOVER
202 and reset_link returns PCI_ERS_RESULT_RECOVERE 202 and reset_link returns PCI_ERS_RESULT_RECOVERED, the error handling goes
203 to mmio_enabled. 203 to mmio_enabled.
204 204
205 Frequent Asked Questions 205 Frequent Asked Questions
206 ------------------------ 206 ------------------------
207 207
208 Q: 208 Q:
209 What happens if a PCIe device driver does no 209 What happens if a PCIe device driver does not provide an
210 error recovery handler (pci_driver->err_hand 210 error recovery handler (pci_driver->err_handler is equal to NULL)?
211 211
212 A: 212 A:
213 The devices attached with the driver won't b 213 The devices attached with the driver won't be recovered. If the
214 error is fatal, kernel will print out warnin 214 error is fatal, kernel will print out warning messages. Please refer
215 to section 3 for more information. 215 to section 3 for more information.
216 216
217 Q: 217 Q:
218 What happens if an upstream port service dri 218 What happens if an upstream port service driver does not provide
219 callback reset_link? 219 callback reset_link?
220 220
221 A: 221 A:
222 Fatal error recovery will fail if the errors 222 Fatal error recovery will fail if the errors are reported by the
223 upstream ports who are attached by the servi 223 upstream ports who are attached by the service driver.
224 224
225 225
226 Software error injection 226 Software error injection
227 ======================== 227 ========================
228 228
229 Debugging PCIe AER error recovery code is quit 229 Debugging PCIe AER error recovery code is quite difficult because it
230 is hard to trigger real hardware errors. Softw 230 is hard to trigger real hardware errors. Software based error
231 injection can be used to fake various kinds of 231 injection can be used to fake various kinds of PCIe errors.
232 232
233 First you should enable PCIe AER software erro 233 First you should enable PCIe AER software error injection in kernel
234 configuration, that is, following item should 234 configuration, that is, following item should be in your .config.
235 235
236 CONFIG_PCIEAER_INJECT=y or CONFIG_PCIEAER_INJE 236 CONFIG_PCIEAER_INJECT=y or CONFIG_PCIEAER_INJECT=m
237 237
238 After reboot with new kernel or insert the mod 238 After reboot with new kernel or insert the module, a device file named
239 /dev/aer_inject should be created. 239 /dev/aer_inject should be created.
240 240
241 Then, you need a user space tool named aer-inj 241 Then, you need a user space tool named aer-inject, which can be gotten
242 from: 242 from:
243 243
244 https://github.com/intel/aer-inject.git 244 https://github.com/intel/aer-inject.git
245 245
246 More information about aer-inject can be found 246 More information about aer-inject can be found in the document in
247 its source code. 247 its source code.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.