1 The PCI Express Advanced Error Reporting Driver Guide HOWTO
2 T. Long Nguyen <tom.l.nguyen@intel.com>
3 Yanmin Zhang <yanmin.zhang@intel.com>
11 This guide describes the basics of the PCI Express Advanced Error
12 Reporting (AER) driver and provides information on how to use it, as
13 well as how to enable the drivers of endpoint devices to conform with
14 PCI Express AER driver.
16 1.2 Copyright © Intel Corporation 2006.
18 1.3 What is the PCI Express AER Driver?
20 PCI Express error signaling can occur on the PCI Express link itself
21 or on behalf of transactions initiated on the link. PCI Express
22 defines two error reporting paradigms: the baseline capability and
23 the Advanced Error Reporting capability. The baseline capability is
24 required of all PCI Express components providing a minimum defined
25 set of error reporting requirements. Advanced Error Reporting
26 capability is implemented with a PCI Express advanced error reporting
27 extended capability structure providing more robust error reporting.
29 The PCI Express AER driver provides the infrastructure to support PCI
30 Express Advanced Error Reporting capability. The PCI Express AER
31 driver provides three basic functions:
33 - Gathers the comprehensive error information if errors occurred.
34 - Reports error to the users.
35 - Performs error recovery actions.
37 AER driver only attaches root ports which support PCI-Express AER
43 2.1 Include the PCI Express AER Root Driver into the Linux Kernel
45 The PCI Express AER Root driver is a Root Port service driver attached
46 to the PCI Express Port Bus driver. If a user wants to use it, the driver
47 has to be compiled. Option CONFIG_PCIEAER supports this capability. It
48 depends on CONFIG_PCIEPORTBUS, so pls. set CONFIG_PCIEPORTBUS=y and
51 2.2 Load PCI Express AER Root Driver
52 There is a case where a system has AER support in BIOS. Enabling the AER
53 Root driver and having AER support in BIOS may result unpredictable
54 behavior. To avoid this conflict, a successful load of the AER Root driver
55 requires ACPI _OSC support in the BIOS to allow the AER Root driver to
56 request for native control of AER. See the PCI FW 3.0 Specification for
57 details regarding OSC usage. Currently, lots of firmwares don't provide
58 _OSC support while they use PCI Express. To support such firmwares,
59 forceload, a parameter of type bool, could enable AER to continue to
60 be initiated although firmwares have no _OSC support. To enable the
61 walkaround, pls. add aerdriver.forceload=y to kernel boot parameter line
62 when booting kernel. Note that forceload=n by default.
64 nosourceid, another parameter of type bool, can be used when broken
65 hardware (mostly chipsets) has root ports that cannot obtain the reporting
66 source ID. nosourceid=n by default.
69 When a PCI-E AER error is captured, an error message will be outputed to
70 console. If it's a correctable error, it is outputed as a warning.
71 Otherwise, it is printed as an error. So users could choose different
72 log level to filter out correctable error messages.
74 Below shows an example.
75 +------ PCI-Express Device Error -----+
76 Error Severity : Uncorrected (Fatal)
77 PCIE Bus Error type : Transaction Layer
78 Unsupported Request : First
80 VendorID=8086h, DeviceID=0329h, Bus=05h, Device=00h, Function=00h
82 04000001 00200a03 05010000 00050100
84 In the example, 'Requester ID' means the ID of the device who sends
85 the error message to root port. Pls. refer to pci express specs for
91 To enable AER aware support requires a software driver to configure
92 the AER capability structure within its device and to provide callbacks.
94 To support AER better, developers need understand how AER does work
97 PCI Express errors are classified into two types: correctable errors
98 and uncorrectable errors. This classification is based on the impacts
99 of those errors, which may result in degraded performance or function
102 Correctable errors pose no impacts on the functionality of the
103 interface. The PCI Express protocol can recover without any software
104 intervention or any loss of data. These errors are detected and
105 corrected by hardware. Unlike correctable errors, uncorrectable
106 errors impact functionality of the interface. Uncorrectable errors
107 can cause a particular transaction or a particular PCI Express link
108 to be unreliable. Depending on those error conditions, uncorrectable
109 errors are further classified into non-fatal errors and fatal errors.
110 Non-fatal errors cause the particular transaction to be unreliable,
111 but the PCI Express link itself is fully functional. Fatal errors, on
112 the other hand, cause the link to be unreliable.
114 When AER is enabled, a PCI Express device will automatically send an
115 error message to the PCIE root port above it when the device captures
116 an error. The Root Port, upon receiving an error reporting message,
117 internally processes and logs the error message in its PCI Express
118 capability structure. Error information being logged includes storing
119 the error reporting agent's requestor ID into the Error Source
120 Identification Registers and setting the error bits of the Root Error
121 Status Register accordingly. If AER error reporting is enabled in Root
122 Error Command Register, the Root Port generates an interrupt if an
125 Note that the errors as described above are related to the PCI Express
126 hierarchy and links. These errors do not include any device specific
127 errors because device specific errors will still get sent directly to
130 3.1 Configure the AER capability structure
132 AER aware drivers of PCI Express component need change the device
133 control registers to enable AER. They also could change AER registers,
134 including mask and severity registers. Helper function
135 pci_enable_pcie_error_reporting could be used to enable AER. See
138 3.2. Provide callbacks
140 3.2.1 callback reset_link to reset pci express link
142 This callback is used to reset the pci express physical link when a
143 fatal error happens. The root port aer service driver provides a
144 default reset_link function, but different upstream ports might
145 have different specifications to reset pci express link, so all
146 upstream ports should provide their own reset_link functions.
148 In struct pcie_port_service_driver, a new pointer, reset_link, is
151 pci_ers_result_t (*reset_link) (struct pci_dev *dev);
153 Section 3.2.2.2 provides more detailed info on when to call
156 3.2.2 PCI error-recovery callbacks
158 The PCI Express AER Root driver uses error callbacks to coordinate
159 with downstream device drivers associated with a hierarchy in question
160 when performing error recovery actions.
162 Data struct pci_driver has a pointer, err_handler, to point to
163 pci_error_handlers who consists of a couple of callback function
164 pointers. AER driver follows the rules defined in
165 pci-error-recovery.txt except pci express specific parts (e.g.
166 reset_link). Pls. refer to pci-error-recovery.txt for detailed
167 definitions of the callbacks.
169 Below sections specify when to call the error callback functions.
171 3.2.2.1 Correctable errors
173 Correctable errors pose no impacts on the functionality of
174 the interface. The PCI Express protocol can recover without any
175 software intervention or any loss of data. These errors do not
176 require any recovery actions. The AER driver clears the device's
177 correctable error status register accordingly and logs these errors.
179 3.2.2.2 Non-correctable (non-fatal and fatal) errors
181 If an error message indicates a non-fatal error, performing link reset
182 at upstream is not required. The AER driver calls error_detected(dev,
183 pci_channel_io_normal) to all drivers associated within a hierarchy in
184 question. for example,
185 EndPoint<==>DownstreamPort B<==>UpstreamPort A<==>RootPort.
186 If Upstream port A captures an AER error, the hierarchy consists of
187 Downstream port B and EndPoint.
189 A driver may return PCI_ERS_RESULT_CAN_RECOVER,
190 PCI_ERS_RESULT_DISCONNECT, or PCI_ERS_RESULT_NEED_RESET, depending on
191 whether it can recover or the AER driver calls mmio_enabled as next.
193 If an error message indicates a fatal error, kernel will broadcast
194 error_detected(dev, pci_channel_io_frozen) to all drivers within
195 a hierarchy in question. Then, performing link reset at upstream is
196 necessary. As different kinds of devices might use different approaches
197 to reset link, AER port service driver is required to provide the
198 function to reset link. Firstly, kernel looks for if the upstream
199 component has an aer driver. If it has, kernel uses the reset_link
200 callback of the aer driver. If the upstream component has no aer driver
201 and the port is downstream port, we will use the aer driver of the
202 root port who reports the AER error. As for upstream ports,
203 they should provide their own aer service drivers with reset_link
204 function. If error_detected returns PCI_ERS_RESULT_CAN_RECOVER and
205 reset_link returns PCI_ERS_RESULT_RECOVERED, the error handling goes
210 3.3.1 int pci_enable_pcie_error_reporting(struct pci_dev *dev);
211 pci_enable_pcie_error_reporting enables the device to send error
212 messages to root port when an error is detected. Note that devices
213 don't enable the error reporting by default, so device drivers need
214 call this function to enable it.
216 3.3.2 int pci_disable_pcie_error_reporting(struct pci_dev *dev);
217 pci_disable_pcie_error_reporting disables the device to send error
218 messages to root port when an error is detected.
220 3.3.3 int pci_cleanup_aer_uncorrect_error_status(struct pci_dev *dev);
221 pci_cleanup_aer_uncorrect_error_status cleanups the uncorrectable
222 error status register.
224 3.4 Frequent Asked Questions
226 Q: What happens if a PCI Express device driver does not provide an
227 error recovery handler (pci_driver->err_handler is equal to NULL)?
229 A: The devices attached with the driver won't be recovered. If the
230 error is fatal, kernel will print out warning messages. Please refer
231 to section 3 for more information.
233 Q: What happens if an upstream port service driver does not provide
236 A: Fatal error recovery will fail if the errors are reported by the
237 upstream ports who are attached by the service driver.
239 Q: How does this infrastructure deal with driver that is not PCI
242 A: This infrastructure calls the error callback functions of the
243 driver when an error happens. But if the driver is not aware of
244 PCI Express, the device might not report its own errors to root
247 Q: What modifications will that driver need to make it compatible
248 with the PCI Express AER Root driver?
250 A: It could call the helper functions to enable AER in devices and
251 cleanup uncorrectable status register. Pls. refer to section 3.3.
254 4. Software error injection
256 Debugging PCIE AER error recovery code is quite difficult because it
257 is hard to trigger real hardware errors. Software based error
258 injection can be used to fake various kinds of PCIE errors.
260 First you should enable PCIE AER software error injection in kernel
261 configuration, that is, following item should be in your .config.
263 CONFIG_PCIEAER_INJECT=y or CONFIG_PCIEAER_INJECT=m
265 After reboot with new kernel or insert the module, a device file named
266 /dev/aer_inject should be created.
268 Then, you need a user space tool named aer-inject, which can be gotten
270 http://www.kernel.org/pub/linux/utils/pci/aer-inject/
272 More information about aer-inject can be found in the document comes
273 with its source code.