1.. SPDX-License-Identifier: (GPL-2.0+ OR CC-BY-4.0) 2.. 3 If you want to distribute this text under CC-BY-4.0 only, please use 'The 4 Linux kernel developers' for author attribution and link this as source: 5 https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/plain/Documentation/admin-guide/reporting-issues.rst 6.. 7 Note: Only the content of this RST file as found in the Linux kernel sources 8 is available under CC-BY-4.0, as versions of this text that were processed 9 (for example by the kernel's build system) might contain content taken from 10 files which use a more restrictive license. 11 12 13Reporting issues 14++++++++++++++++ 15 16 17The short guide (aka TL;DR) 18=========================== 19 20Are you facing a regression with vanilla kernels from the same stable or 21longterm series? One still supported? Then search the `LKML 22<https://lore.kernel.org/lkml/>`_ and the `Linux stable mailing list 23<https://lore.kernel.org/stable/>`_ archives for matching reports to join. If 24you don't find any, install `the latest release from that series 25<https://kernel.org/>`_. If it still shows the issue, report it to the stable 26mailing list (stable@vger.kernel.org) and CC the regressions list 27(regressions@lists.linux.dev); ideally also CC the maintainer and the mailing 28list for the subsystem in question. 29 30In all other cases try your best guess which kernel part might be causing the 31issue. Check the :ref:`MAINTAINERS <maintainers>` file for how its developers 32expect to be told about problems, which most of the time will be by email with a 33mailing list in CC. Check the destination's archives for matching reports; 34search the `LKML <https://lore.kernel.org/lkml/>`_ and the web, too. If you 35don't find any to join, install `the latest mainline kernel 36<https://kernel.org/>`_. If the issue is present there, send a report. 37 38The issue was fixed there, but you would like to see it resolved in a still 39supported stable or longterm series as well? Then install its latest release. 40If it shows the problem, search for the change that fixed it in mainline and 41check if backporting is in the works or was discarded; if it's neither, ask 42those who handled the change for it. 43 44**General remarks**: When installing and testing a kernel as outlined above, 45ensure it's vanilla (IOW: not patched and not using add-on modules). Also make 46sure it's built and running in a healthy environment and not already tainted 47before the issue occurs. 48 49If you are facing multiple issues with the Linux kernel at once, report each 50separately. While writing your report, include all information relevant to the 51issue, like the kernel and the distro used. In case of a regression, CC the 52regressions mailing list (regressions@lists.linux.dev) to your report. Also try 53to pin-point the culprit with a bisection; if you succeed, include its 54commit-id and CC everyone in the sign-off-by chain. 55 56Once the report is out, answer any questions that come up and help where you 57can. That includes keeping the ball rolling by occasionally retesting with newer 58releases and sending a status update afterwards. 59 60Step-by-step guide how to report issues to the kernel maintainers 61================================================================= 62 63The above TL;DR outlines roughly how to report issues to the Linux kernel 64developers. It might be all that's needed for people already familiar with 65reporting issues to Free/Libre & Open Source Software (FLOSS) projects. For 66everyone else there is this section. It is more detailed and uses a 67step-by-step approach. It still tries to be brief for readability and leaves 68out a lot of details; those are described below the step-by-step guide in a 69reference section, which explains each of the steps in more detail. 70 71Note: this section covers a few more aspects than the TL;DR and does things in 72a slightly different order. That's in your interest, to make sure you notice 73early if an issue that looks like a Linux kernel problem is actually caused by 74something else. These steps thus help to ensure the time you invest in this 75process won't feel wasted in the end: 76 77 * Are you facing an issue with a Linux kernel a hardware or software vendor 78 provided? Then in almost all cases you are better off to stop reading this 79 document and reporting the issue to your vendor instead, unless you are 80 willing to install the latest Linux version yourself. Be aware the latter 81 will often be needed anyway to hunt down and fix issues. 82 83 * Perform a rough search for existing reports with your favorite internet 84 search engine; additionally, check the archives of the `Linux Kernel Mailing 85 List (LKML) <https://lore.kernel.org/lkml/>`_. If you find matching reports, 86 join the discussion instead of sending a new one. 87 88 * See if the issue you are dealing with qualifies as regression, security 89 issue, or a really severe problem: those are 'issues of high priority' that 90 need special handling in some steps that are about to follow. 91 92 * Make sure it's not the kernel's surroundings that are causing the issue 93 you face. 94 95 * Create a fresh backup and put system repair and restore tools at hand. 96 97 * Ensure your system does not enhance its kernels by building additional 98 kernel modules on-the-fly, which solutions like DKMS might be doing locally 99 without your knowledge. 100 101 * Check if your kernel was 'tainted' when the issue occurred, as the event 102 that made the kernel set this flag might be causing the issue you face. 103 104 * Write down coarsely how to reproduce the issue. If you deal with multiple 105 issues at once, create separate notes for each of them and make sure they 106 work independently on a freshly booted system. That's needed, as each issue 107 needs to get reported to the kernel developers separately, unless they are 108 strongly entangled. 109 110 * If you are facing a regression within a stable or longterm version line 111 (say something broke when updating from 5.10.4 to 5.10.5), scroll down to 112 'Dealing with regressions within a stable and longterm kernel line'. 113 114 * Locate the driver or kernel subsystem that seems to be causing the issue. 115 Find out how and where its developers expect reports. Note: most of the 116 time this won't be bugzilla.kernel.org, as issues typically need to be sent 117 by mail to a maintainer and a public mailing list. 118 119 * Search the archives of the bug tracker or mailing list in question 120 thoroughly for reports that might match your issue. If you find anything, 121 join the discussion instead of sending a new report. 122 123After these preparations you'll now enter the main part: 124 125 * Unless you are already running the latest 'mainline' Linux kernel, better 126 go and install it for the reporting process. Testing and reporting with 127 the latest 'stable' Linux can be an acceptable alternative in some 128 situations; during the merge window that actually might be even the best 129 approach, but in that development phase it can be an even better idea to 130 suspend your efforts for a few days anyway. Whatever version you choose, 131 ideally use a 'vanilla' build. Ignoring these advices will dramatically 132 increase the risk your report will be rejected or ignored. 133 134 * Ensure the kernel you just installed does not 'taint' itself when 135 running. 136 137 * Reproduce the issue with the kernel you just installed. If it doesn't show 138 up there, scroll down to the instructions for issues only happening with 139 stable and longterm kernels. 140 141 * Optimize your notes: try to find and write the most straightforward way to 142 reproduce your issue. Make sure the end result has all the important 143 details, and at the same time is easy to read and understand for others 144 that hear about it for the first time. And if you learned something in this 145 process, consider searching again for existing reports about the issue. 146 147 * If your failure involves a 'panic', 'Oops', 'warning', or 'BUG', consider 148 decoding the kernel log to find the line of code that triggered the error. 149 150 * If your problem is a regression, try to narrow down when the issue was 151 introduced as much as possible. 152 153 * Start to compile the report by writing a detailed description about the 154 issue. Always mention a few things: the latest kernel version you installed 155 for reproducing, the Linux Distribution used, and your notes on how to 156 reproduce the issue. Ideally, make the kernel's build configuration 157 (.config) and the output from ``dmesg`` available somewhere on the net and 158 link to it. Include or upload all other information that might be relevant, 159 like the output/screenshot of an Oops or the output from ``lspci``. Once 160 you wrote this main part, insert a normal length paragraph on top of it 161 outlining the issue and the impact quickly. On top of this add one sentence 162 that briefly describes the problem and gets people to read on. Now give the 163 thing a descriptive title or subject that yet again is shorter. Then you're 164 ready to send or file the report like the MAINTAINERS file told you, unless 165 you are dealing with one of those 'issues of high priority': they need 166 special care which is explained in 'Special handling for high priority 167 issues' below. 168 169 * Wait for reactions and keep the thing rolling until you can accept the 170 outcome in one way or the other. Thus react publicly and in a timely manner 171 to any inquiries. Test proposed fixes. Do proactive testing: retest with at 172 least every first release candidate (RC) of a new mainline version and 173 report your results. Send friendly reminders if things stall. And try to 174 help yourself, if you don't get any help or if it's unsatisfying. 175 176 177Reporting regressions within a stable and longterm kernel line 178-------------------------------------------------------------- 179 180This subsection is for you, if you followed above process and got sent here at 181the point about regression within a stable or longterm kernel version line. You 182face one of those if something breaks when updating from 5.10.4 to 5.10.5 (a 183switch from 5.9.15 to 5.10.5 does not qualify). The developers want to fix such 184regressions as quickly as possible, hence there is a streamlined process to 185report them: 186 187 * Check if the kernel developers still maintain the Linux kernel version 188 line you care about: go to the `front page of kernel.org 189 <https://kernel.org/>`_ and make sure it mentions 190 the latest release of the particular version line without an '[EOL]' tag. 191 192 * Check the archives of the `Linux stable mailing list 193 <https://lore.kernel.org/stable/>`_ for existing reports. 194 195 * Install the latest release from the particular version line as a vanilla 196 kernel. Ensure this kernel is not tainted and still shows the problem, as 197 the issue might have already been fixed there. If you first noticed the 198 problem with a vendor kernel, check a vanilla build of the last version 199 known to work performs fine as well. 200 201 * Send a short problem report to the Linux stable mailing list 202 (stable@vger.kernel.org) and CC the Linux regressions mailing list 203 (regressions@lists.linux.dev); if you suspect the cause in a particular 204 subsystem, CC its maintainer and its mailing list. Roughly describe the 205 issue and ideally explain how to reproduce it. Mention the first version 206 that shows the problem and the last version that's working fine. Then 207 wait for further instructions. 208 209The reference section below explains each of these steps in more detail. 210 211 212Reporting issues only occurring in older kernel version lines 213------------------------------------------------------------- 214 215This subsection is for you, if you tried the latest mainline kernel as outlined 216above, but failed to reproduce your issue there; at the same time you want to 217see the issue fixed in a still supported stable or longterm series or vendor 218kernels regularly rebased on those. If that the case, follow these steps: 219 220 * Prepare yourself for the possibility that going through the next few steps 221 might not get the issue solved in older releases: the fix might be too big 222 or risky to get backported there. 223 224 * Perform the first three steps in the section "Dealing with regressions 225 within a stable and longterm kernel line" above. 226 227 * Search the Linux kernel version control system for the change that fixed 228 the issue in mainline, as its commit message might tell you if the fix is 229 scheduled for backporting already. If you don't find anything that way, 230 search the appropriate mailing lists for posts that discuss such an issue 231 or peer-review possible fixes; then check the discussions if the fix was 232 deemed unsuitable for backporting. If backporting was not considered at 233 all, join the newest discussion, asking if it's in the cards. 234 235 * One of the former steps should lead to a solution. If that doesn't work 236 out, ask the maintainers for the subsystem that seems to be causing the 237 issue for advice; CC the mailing list for the particular subsystem as well 238 as the stable mailing list. 239 240The reference section below explains each of these steps in more detail. 241 242 243Reference section: Reporting issues to the kernel maintainers 244============================================================= 245 246The detailed guides above outline all the major steps in brief fashion, which 247should be enough for most people. But sometimes there are situations where even 248experienced users might wonder how to actually do one of those steps. That's 249what this section is for, as it will provide a lot more details on each of the 250above steps. Consider this as reference documentation: it's possible to read it 251from top to bottom. But it's mainly meant to skim over and a place to look up 252details how to actually perform those steps. 253 254A few words of general advice before digging into the details: 255 256 * The Linux kernel developers are well aware this process is complicated and 257 demands more than other FLOSS projects. We'd love to make it simpler. But 258 that would require work in various places as well as some infrastructure, 259 which would need constant maintenance; nobody has stepped up to do that 260 work, so that's just how things are for now. 261 262 * A warranty or support contract with some vendor doesn't entitle you to 263 request fixes from developers in the upstream Linux kernel community: such 264 contracts are completely outside the scope of the Linux kernel, its 265 development community, and this document. That's why you can't demand 266 anything such a contract guarantees in this context, not even if the 267 developer handling the issue works for the vendor in question. If you want 268 to claim your rights, use the vendor's support channel instead. When doing 269 so, you might want to mention you'd like to see the issue fixed in the 270 upstream Linux kernel; motivate them by saying it's the only way to ensure 271 the fix in the end will get incorporated in all Linux distributions. 272 273 * If you never reported an issue to a FLOSS project before you should consider 274 reading `How to Report Bugs Effectively 275 <https://www.chiark.greenend.org.uk/~sgtatham/bugs.html>`_, `How To Ask 276 Questions The Smart Way 277 <http://www.catb.org/esr/faqs/smart-questions.html>`_, and `How to ask good 278 questions <https://jvns.ca/blog/good-questions/>`_. 279 280With that off the table, find below the details on how to properly report 281issues to the Linux kernel developers. 282 283 284Make sure you're using the upstream Linux kernel 285------------------------------------------------ 286 287 *Are you facing an issue with a Linux kernel a hardware or software vendor 288 provided? Then in almost all cases you are better off to stop reading this 289 document and reporting the issue to your vendor instead, unless you are 290 willing to install the latest Linux version yourself. Be aware the latter 291 will often be needed anyway to hunt down and fix issues.* 292 293Like most programmers, Linux kernel developers don't like to spend time dealing 294with reports for issues that don't even happen with their current code. It's 295just a waste everybody's time, especially yours. Unfortunately such situations 296easily happen when it comes to the kernel and often leads to frustration on both 297sides. That's because almost all Linux-based kernels pre-installed on devices 298(Computers, Laptops, Smartphones, Routers, …) and most shipped by Linux 299distributors are quite distant from the official Linux kernel as distributed by 300kernel.org: these kernels from these vendors are often ancient from the point of 301Linux development or heavily modified, often both. 302 303Most of these vendor kernels are quite unsuitable for reporting issues to the 304Linux kernel developers: an issue you face with one of them might have been 305fixed by the Linux kernel developers months or years ago already; additionally, 306the modifications and enhancements by the vendor might be causing the issue you 307face, even if they look small or totally unrelated. That's why you should report 308issues with these kernels to the vendor. Its developers should look into the 309report and, in case it turns out to be an upstream issue, fix it directly 310upstream or forward the report there. In practice that often does not work out 311or might not what you want. You thus might want to consider circumventing the 312vendor by installing the very latest Linux kernel core yourself. If that's an 313option for you move ahead in this process, as a later step in this guide will 314explain how to do that once it rules out other potential causes for your issue. 315 316Note, the previous paragraph is starting with the word 'most', as sometimes 317developers in fact are willing to handle reports about issues occurring with 318vendor kernels. If they do in the end highly depends on the developers and the 319issue in question. Your chances are quite good if the distributor applied only 320small modifications to a kernel based on a recent Linux version; that for 321example often holds true for the mainline kernels shipped by Debian GNU/Linux 322Sid or Fedora Rawhide. Some developers will also accept reports about issues 323with kernels from distributions shipping the latest stable kernel, as long as 324its only slightly modified; that for example is often the case for Arch Linux, 325regular Fedora releases, and openSUSE Tumbleweed. But keep in mind, you better 326want to use a mainline Linux and avoid using a stable kernel for this 327process, as outlined in the section 'Install a fresh kernel for testing' in more 328detail. 329 330Obviously you are free to ignore all this advice and report problems with an old 331or heavily modified vendor kernel to the upstream Linux developers. But note, 332those often get rejected or ignored, so consider yourself warned. But it's still 333better than not reporting the issue at all: sometimes such reports directly or 334indirectly will help to get the issue fixed over time. 335 336 337Search for existing reports, first run 338-------------------------------------- 339 340 *Perform a rough search for existing reports with your favorite internet 341 search engine; additionally, check the archives of the Linux Kernel Mailing 342 List (LKML). If you find matching reports, join the discussion instead of 343 sending a new one.* 344 345Reporting an issue that someone else already brought forward is often a waste of 346time for everyone involved, especially you as the reporter. So it's in your own 347interest to thoroughly check if somebody reported the issue already. At this 348step of the process it's okay to just perform a rough search: a later step will 349tell you to perform a more detailed search once you know where your issue needs 350to be reported to. Nevertheless, do not hurry with this step of the reporting 351process, it can save you time and trouble. 352 353Simply search the internet with your favorite search engine first. Afterwards, 354search the `Linux Kernel Mailing List (LKML) archives 355<https://lore.kernel.org/lkml/>`_. 356 357If you get flooded with results consider telling your search engine to limit 358search timeframe to the past month or year. And wherever you search, make sure 359to use good search terms; vary them a few times, too. While doing so try to 360look at the issue from the perspective of someone else: that will help you to 361come up with other words to use as search terms. Also make sure not to use too 362many search terms at once. Remember to search with and without information like 363the name of the kernel driver or the name of the affected hardware component. 364But its exact brand name (say 'ASUS Red Devil Radeon RX 5700 XT Gaming OC') 365often is not much helpful, as it is too specific. Instead try search terms like 366the model line (Radeon 5700 or Radeon 5000) and the code name of the main chip 367('Navi' or 'Navi10') with and without its manufacturer ('AMD'). 368 369In case you find an existing report about your issue, join the discussion, as 370you might be able to provide valuable additional information. That can be 371important even when a fix is prepared or in its final stages already, as 372developers might look for people that can provide additional information or 373test a proposed fix. Jump to the section 'Duties after the report went out' for 374details on how to get properly involved. 375 376Note, searching `bugzilla.kernel.org <https://bugzilla.kernel.org/>`_ might also 377be a good idea, as that might provide valuable insights or turn up matching 378reports. If you find the latter, just keep in mind: most subsystems expect 379reports in different places, as described below in the section "Check where you 380need to report your issue". The developers that should take care of the issue 381thus might not even be aware of the bugzilla ticket. Hence, check the ticket if 382the issue already got reported as outlined in this document and if not consider 383doing so. 384 385 386Issue of high priority? 387----------------------- 388 389 *See if the issue you are dealing with qualifies as regression, security 390 issue, or a really severe problem: those are 'issues of high priority' that 391 need special handling in some steps that are about to follow.* 392 393Linus Torvalds and the leading Linux kernel developers want to see some issues 394fixed as soon as possible, hence there are 'issues of high priority' that get 395handled slightly differently in the reporting process. Three type of cases 396qualify: regressions, security issues, and really severe problems. 397 398You deal with a 'regression' if something that worked with an older version of 399the Linux kernel does not work with a newer one or somehow works worse with it. 400It thus is a regression when a WiFi driver that did a fine job with Linux 5.7 401somehow misbehaves with 5.8 or doesn't work at all. It's also a regression if 402an application shows erratic behavior with a newer kernel, which might happen 403due to incompatible changes in the interface between the kernel and the 404userland (like procfs and sysfs). Significantly reduced performance or 405increased power consumption also qualify as regression. But keep in mind: the 406new kernel needs to be built with a configuration that is similar to the one 407from the old kernel (see below how to achieve that). That's because the kernel 408developers sometimes can not avoid incompatibilities when implementing new 409features; but to avoid regressions such features have to be enabled explicitly 410during build time configuration. 411 412What qualifies as security issue is left to your judgment. Consider reading 413'Documentation/admin-guide/security-bugs.rst' before proceeding, as it 414provides additional details how to best handle security issues. 415 416An issue is a 'really severe problem' when something totally unacceptably bad 417happens. That's for example the case when a Linux kernel corrupts the data it's 418handling or damages hardware it's running on. You're also dealing with a severe 419issue when the kernel suddenly stops working with an error message ('kernel 420panic') or without any farewell note at all. Note: do not confuse a 'panic' (a 421fatal error where the kernel stop itself) with a 'Oops' (a recoverable error), 422as the kernel remains running after the latter. 423 424 425Ensure a healthy environment 426---------------------------- 427 428 *Make sure it's not the kernel's surroundings that are causing the issue 429 you face.* 430 431Problems that look a lot like a kernel issue are sometimes caused by build or 432runtime environment. It's hard to rule out that problem completely, but you 433should minimize it: 434 435 * Use proven tools when building your kernel, as bugs in the compiler or the 436 binutils can cause the resulting kernel to misbehave. 437 438 * Ensure your computer components run within their design specifications; 439 that's especially important for the main processor, the main memory, and the 440 motherboard. Therefore, stop undervolting or overclocking when facing a 441 potential kernel issue. 442 443 * Try to make sure it's not faulty hardware that is causing your issue. Bad 444 main memory for example can result in a multitude of issues that will 445 manifest itself in problems looking like kernel issues. 446 447 * If you're dealing with a filesystem issue, you might want to check the file 448 system in question with ``fsck``, as it might be damaged in a way that leads 449 to unexpected kernel behavior. 450 451 * When dealing with a regression, make sure it's not something else that 452 changed in parallel to updating the kernel. The problem for example might be 453 caused by other software that was updated at the same time. It can also 454 happen that a hardware component coincidentally just broke when you rebooted 455 into a new kernel for the first time. Updating the systems BIOS or changing 456 something in the BIOS Setup can also lead to problems that on look a lot 457 like a kernel regression. 458 459 460Prepare for emergencies 461----------------------- 462 463 *Create a fresh backup and put system repair and restore tools at hand.* 464 465Reminder, you are dealing with computers, which sometimes do unexpected things, 466especially if you fiddle with crucial parts like the kernel of its operating 467system. That's what you are about to do in this process. Thus, make sure to 468create a fresh backup; also ensure you have all tools at hand to repair or 469reinstall the operating system as well as everything you need to restore the 470backup. 471 472 473Make sure your kernel doesn't get enhanced 474------------------------------------------ 475 476 *Ensure your system does not enhance its kernels by building additional 477 kernel modules on-the-fly, which solutions like DKMS might be doing locally 478 without your knowledge.* 479 480The risk your issue report gets ignored or rejected dramatically increases if 481your kernel gets enhanced in any way. That's why you should remove or disable 482mechanisms like akmods and DKMS: those build add-on kernel modules 483automatically, for example when you install a new Linux kernel or boot it for 484the first time. Also remove any modules they might have installed. Then reboot 485before proceeding. 486 487Note, you might not be aware that your system is using one of these solutions: 488they often get set up silently when you install Nvidia's proprietary graphics 489driver, VirtualBox, or other software that requires a some support from a 490module not part of the Linux kernel. That why your might need to uninstall the 491packages with such software to get rid of any 3rd party kernel module. 492 493 494Check 'taint' flag 495------------------ 496 497 *Check if your kernel was 'tainted' when the issue occurred, as the event 498 that made the kernel set this flag might be causing the issue you face.* 499 500The kernel marks itself with a 'taint' flag when something happens that might 501lead to follow-up errors that look totally unrelated. The issue you face might 502be such an error if your kernel is tainted. That's why it's in your interest to 503rule this out early before investing more time into this process. This is the 504only reason why this step is here, as this process later will tell you to 505install the latest mainline kernel; you will need to check the taint flag again 506then, as that's when it matters because it's the kernel the report will focus 507on. 508 509On a running system is easy to check if the kernel tainted itself: if ``cat 510/proc/sys/kernel/tainted`` returns '0' then the kernel is not tainted and 511everything is fine. Checking that file is impossible in some situations; that's 512why the kernel also mentions the taint status when it reports an internal 513problem (a 'kernel bug'), a recoverable error (a 'kernel Oops') or a 514non-recoverable error before halting operation (a 'kernel panic'). Look near 515the top of the error messages printed when one of these occurs and search for a 516line starting with 'CPU:'. It should end with 'Not tainted' if the kernel was 517not tainted when it noticed the problem; it was tainted if you see 'Tainted:' 518followed by a few spaces and some letters. 519 520If your kernel is tainted, study 'Documentation/admin-guide/tainted-kernels.rst' 521to find out why. Try to eliminate the reason. Often it's caused by one these 522three things: 523 524 1. A recoverable error (a 'kernel Oops') occurred and the kernel tainted 525 itself, as the kernel knows it might misbehave in strange ways after that 526 point. In that case check your kernel or system log and look for a section 527 that starts with this:: 528 529 Oops: 0000 [#1] SMP 530 531 That's the first Oops since boot-up, as the '#1' between the brackets shows. 532 Every Oops and any other problem that happens after that point might be a 533 follow-up problem to that first Oops, even if both look totally unrelated. 534 Rule this out by getting rid of the cause for the first Oops and reproducing 535 the issue afterwards. Sometimes simply restarting will be enough, sometimes 536 a change to the configuration followed by a reboot can eliminate the Oops. 537 But don't invest too much time into this at this point of the process, as 538 the cause for the Oops might already be fixed in the newer Linux kernel 539 version you are going to install later in this process. 540 541 2. Your system uses a software that installs its own kernel modules, for 542 example Nvidia's proprietary graphics driver or VirtualBox. The kernel 543 taints itself when it loads such module from external sources (even if 544 they are Open Source): they sometimes cause errors in unrelated kernel 545 areas and thus might be causing the issue you face. You therefore have to 546 prevent those modules from loading when you want to report an issue to the 547 Linux kernel developers. Most of the time the easiest way to do that is: 548 temporarily uninstall such software including any modules they might have 549 installed. Afterwards reboot. 550 551 3. The kernel also taints itself when it's loading a module that resides in 552 the staging tree of the Linux kernel source. That's a special area for 553 code (mostly drivers) that does not yet fulfill the normal Linux kernel 554 quality standards. When you report an issue with such a module it's 555 obviously okay if the kernel is tainted; just make sure the module in 556 question is the only reason for the taint. If the issue happens in an 557 unrelated area reboot and temporarily block the module from being loaded 558 by specifying ``foo.blacklist=1`` as kernel parameter (replace 'foo' with 559 the name of the module in question). 560 561 562Document how to reproduce issue 563------------------------------- 564 565 *Write down coarsely how to reproduce the issue. If you deal with multiple 566 issues at once, create separate notes for each of them and make sure they 567 work independently on a freshly booted system. That's needed, as each issue 568 needs to get reported to the kernel developers separately, unless they are 569 strongly entangled.* 570 571If you deal with multiple issues at once, you'll have to report each of them 572separately, as they might be handled by different developers. Describing 573various issues in one report also makes it quite difficult for others to tear 574it apart. Hence, only combine issues in one report if they are very strongly 575entangled. 576 577Additionally, during the reporting process you will have to test if the issue 578happens with other kernel versions. Therefore, it will make your work easier if 579you know exactly how to reproduce an issue quickly on a freshly booted system. 580 581Note: it's often fruitless to report issues that only happened once, as they 582might be caused by a bit flip due to cosmic radiation. That's why you should 583try to rule that out by reproducing the issue before going further. Feel free 584to ignore this advice if you are experienced enough to tell a one-time error 585due to faulty hardware apart from a kernel issue that rarely happens and thus 586is hard to reproduce. 587 588 589Regression in stable or longterm kernel? 590---------------------------------------- 591 592 *If you are facing a regression within a stable or longterm version line 593 (say something broke when updating from 5.10.4 to 5.10.5), scroll down to 594 'Dealing with regressions within a stable and longterm kernel line'.* 595 596Regression within a stable and longterm kernel version line are something the 597Linux developers want to fix badly, as such issues are even more unwanted than 598regression in the main development branch, as they can quickly affect a lot of 599people. The developers thus want to learn about such issues as quickly as 600possible, hence there is a streamlined process to report them. Note, 601regressions with newer kernel version line (say something broke when switching 602from 5.9.15 to 5.10.5) do not qualify. 603 604 605Check where you need to report your issue 606----------------------------------------- 607 608 *Locate the driver or kernel subsystem that seems to be causing the issue. 609 Find out how and where its developers expect reports. Note: most of the 610 time this won't be bugzilla.kernel.org, as issues typically need to be sent 611 by mail to a maintainer and a public mailing list.* 612 613It's crucial to send your report to the right people, as the Linux kernel is a 614big project and most of its developers are only familiar with a small subset of 615it. Quite a few programmers for example only care for just one driver, for 616example one for a WiFi chip; its developer likely will only have small or no 617knowledge about the internals of remote or unrelated "subsystems", like the TCP 618stack, the PCIe/PCI subsystem, memory management or file systems. 619 620Problem is: the Linux kernel lacks a central bug tracker where you can simply 621file your issue and make it reach the developers that need to know about it. 622That's why you have to find the right place and way to report issues yourself. 623You can do that with the help of a script (see below), but it mainly targets 624kernel developers and experts. For everybody else the MAINTAINERS file is the 625better place. 626 627How to read the MAINTAINERS file 628~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 629To illustrate how to use the :ref:`MAINTAINERS <maintainers>` file, lets assume 630the WiFi in your Laptop suddenly misbehaves after updating the kernel. In that 631case it's likely an issue in the WiFi driver. Obviously it could also be some 632code it builds upon, but unless you suspect something like that stick to the 633driver. If it's really something else, the driver's developers will get the 634right people involved. 635 636Sadly, there is no way to check which code is driving a particular hardware 637component that is both universal and easy. 638 639In case of a problem with the WiFi driver you for example might want to look at 640the output of ``lspci -k``, as it lists devices on the PCI/PCIe bus and the 641kernel module driving it:: 642 643 [user@something ~]$ lspci -k 644 [...] 645 3a:00.0 Network controller: Qualcomm Atheros QCA6174 802.11ac Wireless Network Adapter (rev 32) 646 Subsystem: Bigfoot Networks, Inc. Device 1535 647 Kernel driver in use: ath10k_pci 648 Kernel modules: ath10k_pci 649 [...] 650 651But this approach won't work if your WiFi chip is connected over USB or some 652other internal bus. In those cases you might want to check your WiFi manager or 653the output of ``ip link``. Look for the name of the problematic network 654interface, which might be something like 'wlp58s0'. This name can be used like 655this to find the module driving it:: 656 657 [user@something ~]$ realpath --relative-to=/sys/module/ /sys/class/net/wlp58s0/device/driver/module 658 ath10k_pci 659 660In case tricks like these don't bring you any further, try to search the 661internet on how to narrow down the driver or subsystem in question. And if you 662are unsure which it is: just try your best guess, somebody will help you if you 663guessed poorly. 664 665Once you know the driver or subsystem, you want to search for it in the 666MAINTAINERS file. In the case of 'ath10k_pci' you won't find anything, as the 667name is too specific. Sometimes you will need to search on the net for help; 668but before doing so, try a somewhat shorted or modified name when searching the 669MAINTAINERS file, as then you might find something like this:: 670 671 QUALCOMM ATHEROS ATH10K WIRELESS DRIVER 672 Mail: A. Some Human <shuman@example.com> 673 Mailing list: ath10k@lists.infradead.org 674 Status: Supported 675 Web-page: https://wireless.wiki.kernel.org/en/users/Drivers/ath10k 676 SCM: git git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/ath.git 677 Files: drivers/net/wireless/ath/ath10k/ 678 679Note: the line description will be abbreviations, if you read the plain 680MAINTAINERS file found in the root of the Linux source tree. 'Mail:' for 681example will be 'M:', 'Mailing list:' will be 'L', and 'Status:' will be 'S:'. 682A section near the top of the file explains these and other abbreviations. 683 684First look at the line 'Status'. Ideally it should be 'Supported' or 685'Maintained'. If it states 'Obsolete' then you are using some outdated approach 686that was replaced by a newer solution you need to switch to. Sometimes the code 687only has someone who provides 'Odd Fixes' when feeling motivated. And with 688'Orphan' you are totally out of luck, as nobody takes care of the code anymore. 689That only leaves these options: arrange yourself to live with the issue, fix it 690yourself, or find a programmer somewhere willing to fix it. 691 692After checking the status, look for a line starting with 'bugs:': it will tell 693you where to find a subsystem specific bug tracker to file your issue. The 694example above does not have such a line. That is the case for most sections, as 695Linux kernel development is completely driven by mail. Very few subsystems use 696a bug tracker, and only some of those rely on bugzilla.kernel.org. 697 698In this and many other cases you thus have to look for lines starting with 699'Mail:' instead. Those mention the name and the email addresses for the 700maintainers of the particular code. Also look for a line starting with 'Mailing 701list:', which tells you the public mailing list where the code is developed. 702Your report later needs to go by mail to those addresses. Additionally, for all 703issue reports sent by email, make sure to add the Linux Kernel Mailing List 704(LKML) <linux-kernel@vger.kernel.org> to CC. Don't omit either of the mailing 705lists when sending your issue report by mail later! Maintainers are busy people 706and might leave some work for other developers on the subsystem specific list; 707and LKML is important to have one place where all issue reports can be found. 708 709 710Finding the maintainers with the help of a script 711~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 712 713For people that have the Linux sources at hand there is a second option to find 714the proper place to report: the script 'scripts/get_maintainer.pl' which tries 715to find all people to contact. It queries the MAINTAINERS file and needs to be 716called with a path to the source code in question. For drivers compiled as 717module if often can be found with a command like this:: 718 719 $ modinfo ath10k_pci | grep filename | sed 's!/lib/modules/.*/kernel/!!; s!filename:!!; s!\.ko\(\|\.xz\)!!' 720 drivers/net/wireless/ath/ath10k/ath10k_pci.ko 721 722Pass parts of this to the script:: 723 724 $ ./scripts/get_maintainer.pl -f drivers/net/wireless/ath/ath10k* 725 Some Human <shuman@example.com> (supporter:QUALCOMM ATHEROS ATH10K WIRELESS DRIVER) 726 Another S. Human <asomehuman@example.com> (maintainer:NETWORKING DRIVERS) 727 ath10k@lists.infradead.org (open list:QUALCOMM ATHEROS ATH10K WIRELESS DRIVER) 728 linux-wireless@vger.kernel.org (open list:NETWORKING DRIVERS (WIRELESS)) 729 netdev@vger.kernel.org (open list:NETWORKING DRIVERS) 730 linux-kernel@vger.kernel.org (open list) 731 732Don't sent your report to all of them. Send it to the maintainers, which the 733script calls "supporter:"; additionally CC the most specific mailing list for 734the code as well as the Linux Kernel Mailing List (LKML). In this case you thus 735would need to send the report to 'Some Human <shuman@example.com>' with 736'ath10k@lists.infradead.org' and 'linux-kernel@vger.kernel.org' in CC. 737 738Note: in case you cloned the Linux sources with git you might want to call 739``get_maintainer.pl`` a second time with ``--git``. The script then will look 740at the commit history to find which people recently worked on the code in 741question, as they might be able to help. But use these results with care, as it 742can easily send you in a wrong direction. That for example happens quickly in 743areas rarely changed (like old or unmaintained drivers): sometimes such code is 744modified during tree-wide cleanups by developers that do not care about the 745particular driver at all. 746 747 748Search for existing reports, second run 749--------------------------------------- 750 751 *Search the archives of the bug tracker or mailing list in question 752 thoroughly for reports that might match your issue. If you find anything, 753 join the discussion instead of sending a new report.* 754 755As mentioned earlier already: reporting an issue that someone else already 756brought forward is often a waste of time for everyone involved, especially you 757as the reporter. That's why you should search for existing report again, now 758that you know where they need to be reported to. If it's mailing list, you will 759often find its archives on `lore.kernel.org <https://lore.kernel.org/>`_. 760 761But some list are hosted in different places. That for example is the case for 762the ath10k WiFi driver used as example in the previous step. But you'll often 763find the archives for these lists easily on the net. Searching for 'archive 764ath10k@lists.infradead.org' for example will lead you to the `Info page for the 765ath10k mailing list <https://lists.infradead.org/mailman/listinfo/ath10k>`_, 766which at the top links to its 767`list archives <https://lists.infradead.org/pipermail/ath10k/>`_. Sadly this and 768quite a few other lists miss a way to search the archives. In those cases use a 769regular internet search engine and add something like 770'site:lists.infradead.org/pipermail/ath10k/' to your search terms, which limits 771the results to the archives at that URL. 772 773It's also wise to check the internet, LKML and maybe bugzilla.kernel.org again 774at this point. If your report needs to be filed in a bug tracker, you may want 775to check the mailing list archives for the subsystem as well, as someone might 776have reported it only there. 777 778For details how to search and what to do if you find matching reports see 779"Search for existing reports, first run" above. 780 781Do not hurry with this step of the reporting process: spending 30 to 60 minutes 782or even more time can save you and others quite a lot of time and trouble. 783 784 785Install a fresh kernel for testing 786---------------------------------- 787 788 *Unless you are already running the latest 'mainline' Linux kernel, better 789 go and install it for the reporting process. Testing and reporting with 790 the latest 'stable' Linux can be an acceptable alternative in some 791 situations; during the merge window that actually might be even the best 792 approach, but in that development phase it can be an even better idea to 793 suspend your efforts for a few days anyway. Whatever version you choose, 794 ideally use a 'vanilla' built. Ignoring these advices will dramatically 795 increase the risk your report will be rejected or ignored.* 796 797As mentioned in the detailed explanation for the first step already: Like most 798programmers, Linux kernel developers don't like to spend time dealing with 799reports for issues that don't even happen with the current code. It's just a 800waste everybody's time, especially yours. That's why it's in everybody's 801interest that you confirm the issue still exists with the latest upstream code 802before reporting it. You are free to ignore this advice, but as outlined 803earlier: doing so dramatically increases the risk that your issue report might 804get rejected or simply ignored. 805 806In the scope of the kernel "latest upstream" normally means: 807 808 * Install a mainline kernel; the latest stable kernel can be an option, but 809 most of the time is better avoided. Longterm kernels (sometimes called 'LTS 810 kernels') are unsuitable at this point of the process. The next subsection 811 explains all of this in more detail. 812 813 * The over next subsection describes way to obtain and install such a kernel. 814 It also outlines that using a pre-compiled kernel are fine, but better are 815 vanilla, which means: it was built using Linux sources taken straight `from 816 kernel.org <https://kernel.org/>`_ and not modified or enhanced in any way. 817 818Choosing the right version for testing 819~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 820 821Head over to `kernel.org <https://kernel.org/>`_ to find out which version you 822want to use for testing. Ignore the big yellow button that says 'Latest release' 823and look a little lower at the table. At its top you'll see a line starting with 824mainline, which most of the time will point to a pre-release with a version 825number like '5.8-rc2'. If that's the case, you'll want to use this mainline 826kernel for testing, as that where all fixes have to be applied first. Do not let 827that 'rc' scare you, these 'development kernels' are pretty reliable — and you 828made a backup, as you were instructed above, didn't you? 829 830In about two out of every nine to ten weeks, mainline might point you to a 831proper release with a version number like '5.7'. If that happens, consider 832suspending the reporting process until the first pre-release of the next 833version (5.8-rc1) shows up on kernel.org. That's because the Linux development 834cycle then is in its two-week long 'merge window'. The bulk of the changes and 835all intrusive ones get merged for the next release during this time. It's a bit 836more risky to use mainline during this period. Kernel developers are also often 837quite busy then and might have no spare time to deal with issue reports. It's 838also quite possible that one of the many changes applied during the merge 839window fixes the issue you face; that's why you soon would have to retest with 840a newer kernel version anyway, as outlined below in the section 'Duties after 841the report went out'. 842 843That's why it might make sense to wait till the merge window is over. But don't 844to that if you're dealing with something that shouldn't wait. In that case 845consider obtaining the latest mainline kernel via git (see below) or use the 846latest stable version offered on kernel.org. Using that is also acceptable in 847case mainline for some reason does currently not work for you. An in general: 848using it for reproducing the issue is also better than not reporting it issue 849at all. 850 851Better avoid using the latest stable kernel outside merge windows, as all fixes 852must be applied to mainline first. That's why checking the latest mainline 853kernel is so important: any issue you want to see fixed in older version lines 854needs to be fixed in mainline first before it can get backported, which can 855take a few days or weeks. Another reason: the fix you hope for might be too 856hard or risky for backporting; reporting the issue again hence is unlikely to 857change anything. 858 859These aspects are also why longterm kernels (sometimes called "LTS kernels") 860are unsuitable for this part of the reporting process: they are to distant from 861the current code. Hence go and test mainline first and follow the process 862further: if the issue doesn't occur with mainline it will guide you how to get 863it fixed in older version lines, if that's in the cards for the fix in question. 864 865How to obtain a fresh Linux kernel 866~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 867 868**Using a pre-compiled kernel**: This is often the quickest, easiest, and safest 869way for testing — especially is you are unfamiliar with the Linux kernel. The 870problem: most of those shipped by distributors or add-on repositories are build 871from modified Linux sources. They are thus not vanilla and therefore often 872unsuitable for testing and issue reporting: the changes might cause the issue 873you face or influence it somehow. 874 875But you are in luck if you are using a popular Linux distribution: for quite a 876few of them you'll find repositories on the net that contain packages with the 877latest mainline or stable Linux built as vanilla kernel. It's totally okay to 878use these, just make sure from the repository's description they are vanilla or 879at least close to it. Additionally ensure the packages contain the latest 880versions as offered on kernel.org. The packages are likely unsuitable if they 881are older than a week, as new mainline and stable kernels typically get released 882at least once a week. 883 884Please note that you might need to build your own kernel manually later: that's 885sometimes needed for debugging or testing fixes, as described later in this 886document. Also be aware that pre-compiled kernels might lack debug symbols that 887are needed to decode messages the kernel prints when a panic, Oops, warning, or 888BUG occurs; if you plan to decode those, you might be better off compiling a 889kernel yourself (see the end of this subsection and the section titled 'Decode 890failure messages' for details). 891 892**Using git**: Developers and experienced Linux users familiar with git are 893often best served by obtaining the latest Linux kernel sources straight from the 894`official development repository on kernel.org 895<https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/>`_. 896Those are likely a bit ahead of the latest mainline pre-release. Don't worry 897about it: they are as reliable as a proper pre-release, unless the kernel's 898development cycle is currently in the middle of a merge window. But even then 899they are quite reliable. 900 901**Conventional**: People unfamiliar with git are often best served by 902downloading the sources as tarball from `kernel.org <https://kernel.org/>`_. 903 904How to actually build a kernel is not described here, as many websites explain 905the necessary steps already. If you are new to it, consider following one of 906those how-to's that suggest to use ``make localmodconfig``, as that tries to 907pick up the configuration of your current kernel and then tries to adjust it 908somewhat for your system. That does not make the resulting kernel any better, 909but quicker to compile. 910 911Note: If you are dealing with a panic, Oops, warning, or BUG from the kernel, 912please try to enable CONFIG_KALLSYMS when configuring your kernel. 913Additionally, enable CONFIG_DEBUG_KERNEL and CONFIG_DEBUG_INFO, too; the 914latter is the relevant one of those two, but can only be reached if you enable 915the former. Be aware CONFIG_DEBUG_INFO increases the storage space required to 916build a kernel by quite a bit. But that's worth it, as these options will allow 917you later to pinpoint the exact line of code that triggers your issue. The 918section 'Decode failure messages' below explains this in more detail. 919 920But keep in mind: Always keep a record of the issue encountered in case it is 921hard to reproduce. Sending an undecoded report is better than not reporting 922the issue at all. 923 924 925Check 'taint' flag 926------------------ 927 928 *Ensure the kernel you just installed does not 'taint' itself when 929 running.* 930 931As outlined above in more detail already: the kernel sets a 'taint' flag when 932something happens that can lead to follow-up errors that look totally 933unrelated. That's why you need to check if the kernel you just installed does 934not set this flag. And if it does, you in almost all the cases needs to 935eliminate the reason for it before you reporting issues that occur with it. See 936the section above for details how to do that. 937 938 939Reproduce issue with the fresh kernel 940------------------------------------- 941 942 *Reproduce the issue with the kernel you just installed. If it doesn't show 943 up there, scroll down to the instructions for issues only happening with 944 stable and longterm kernels.* 945 946Check if the issue occurs with the fresh Linux kernel version you just 947installed. If it was fixed there already, consider sticking with this version 948line and abandoning your plan to report the issue. But keep in mind that other 949users might still be plagued by it, as long as it's not fixed in either stable 950and longterm version from kernel.org (and thus vendor kernels derived from 951those). If you prefer to use one of those or just want to help their users, 952head over to the section "Details about reporting issues only occurring in 953older kernel version lines" below. 954 955 956Optimize description to reproduce issue 957--------------------------------------- 958 959 *Optimize your notes: try to find and write the most straightforward way to 960 reproduce your issue. Make sure the end result has all the important 961 details, and at the same time is easy to read and understand for others 962 that hear about it for the first time. And if you learned something in this 963 process, consider searching again for existing reports about the issue.* 964 965An unnecessarily complex report will make it hard for others to understand your 966report. Thus try to find a reproducer that's straight forward to describe and 967thus easy to understand in written form. Include all important details, but at 968the same time try to keep it as short as possible. 969 970In this in the previous steps you likely have learned a thing or two about the 971issue you face. Use this knowledge and search again for existing reports 972instead you can join. 973 974 975Decode failure messages 976----------------------- 977 978 *If your failure involves a 'panic', 'Oops', 'warning', or 'BUG', consider 979 decoding the kernel log to find the line of code that triggered the error.* 980 981When the kernel detects an internal problem, it will log some information about 982the executed code. This makes it possible to pinpoint the exact line in the 983source code that triggered the issue and shows how it was called. But that only 984works if you enabled CONFIG_DEBUG_INFO and CONFIG_KALLSYMS when configuring 985your kernel. If you did so, consider to decode the information from the 986kernel's log. That will make it a lot easier to understand what lead to the 987'panic', 'Oops', 'warning', or 'BUG', which increases the chances that someone 988can provide a fix. 989 990Decoding can be done with a script you find in the Linux source tree. If you 991are running a kernel you compiled yourself earlier, call it like this:: 992 993 [user@something ~]$ sudo dmesg | ./linux-5.10.5/scripts/decode_stacktrace.sh ./linux-5.10.5/vmlinux 994 995If you are running a packaged vanilla kernel, you will likely have to install 996the corresponding packages with debug symbols. Then call the script (which you 997might need to get from the Linux sources if your distro does not package it) 998like this:: 999 1000 [user@something ~]$ sudo dmesg | ./linux-5.10.5/scripts/decode_stacktrace.sh \ 1001 /usr/lib/debug/lib/modules/5.10.10-4.1.x86_64/vmlinux /usr/src/kernels/5.10.10-4.1.x86_64/ 1002 1003The script will work on log lines like the following, which show the address of 1004the code the kernel was executing when the error occurred:: 1005 1006 [ 68.387301] RIP: 0010:test_module_init+0x5/0xffa [test_module] 1007 1008Once decoded, these lines will look like this:: 1009 1010 [ 68.387301] RIP: 0010:test_module_init (/home/username/linux-5.10.5/test-module/test-module.c:16) test_module 1011 1012In this case the executed code was built from the file 1013'~/linux-5.10.5/test-module/test-module.c' and the error occurred by the 1014instructions found in line '16'. 1015 1016The script will similarly decode the addresses mentioned in the section 1017starting with 'Call trace', which show the path to the function where the 1018problem occurred. Additionally, the script will show the assembler output for 1019the code section the kernel was executing. 1020 1021Note, if you can't get this to work, simply skip this step and mention the 1022reason for it in the report. If you're lucky, it might not be needed. And if it 1023is, someone might help you to get things going. Also be aware this is just one 1024of several ways to decode kernel stack traces. Sometimes different steps will 1025be required to retrieve the relevant details. Don't worry about that, if that's 1026needed in your case, developers will tell you what to do. 1027 1028 1029Special care for regressions 1030---------------------------- 1031 1032 *If your problem is a regression, try to narrow down when the issue was 1033 introduced as much as possible.* 1034 1035Linux lead developer Linus Torvalds insists that the Linux kernel never 1036worsens, that's why he deems regressions as unacceptable and wants to see them 1037fixed quickly. That's why changes that introduced a regression are often 1038promptly reverted if the issue they cause can't get solved quickly any other 1039way. Reporting a regression is thus a bit like playing a kind of trump card to 1040get something quickly fixed. But for that to happen the change that's causing 1041the regression needs to be known. Normally it's up to the reporter to track 1042down the culprit, as maintainers often won't have the time or setup at hand to 1043reproduce it themselves. 1044 1045To find the change there is a process called 'bisection' which the document 1046'Documentation/admin-guide/bug-bisect.rst' describes in detail. That process 1047will often require you to build about ten to twenty kernel images, trying to 1048reproduce the issue with each of them before building the next. Yes, that takes 1049some time, but don't worry, it works a lot quicker than most people assume. 1050Thanks to a 'binary search' this will lead you to the one commit in the source 1051code management system that's causing the regression. Once you find it, search 1052the net for the subject of the change, its commit id and the shortened commit id 1053(the first 12 characters of the commit id). This will lead you to existing 1054reports about it, if there are any. 1055 1056Note, a bisection needs a bit of know-how, which not everyone has, and quite a 1057bit of effort, which not everyone is willing to invest. Nevertheless, it's 1058highly recommended performing a bisection yourself. If you really can't or 1059don't want to go down that route at least find out which mainline kernel 1060introduced the regression. If something for example breaks when switching from 10615.5.15 to 5.8.4, then try at least all the mainline releases in that area (5.6, 10625.7 and 5.8) to check when it first showed up. Unless you're trying to find a 1063regression in a stable or longterm kernel, avoid testing versions which number 1064has three sections (5.6.12, 5.7.8), as that makes the outcome hard to 1065interpret, which might render your testing useless. Once you found the major 1066version which introduced the regression, feel free to move on in the reporting 1067process. But keep in mind: it depends on the issue at hand if the developers 1068will be able to help without knowing the culprit. Sometimes they might 1069recognize from the report want went wrong and can fix it; other times they will 1070be unable to help unless you perform a bisection. 1071 1072When dealing with regressions make sure the issue you face is really caused by 1073the kernel and not by something else, as outlined above already. 1074 1075In the whole process keep in mind: an issue only qualifies as regression if the 1076older and the newer kernel got built with a similar configuration. The best way 1077to archive this: copy the configuration file (``.config``) from the old working 1078kernel freshly to each newer kernel version you try. Afterwards run ``make 1079olddefconfig`` to adjust it for the needs of the new version. 1080 1081 1082Write and send the report 1083------------------------- 1084 1085 *Start to compile the report by writing a detailed description about the 1086 issue. Always mention a few things: the latest kernel version you installed 1087 for reproducing, the Linux Distribution used, and your notes on how to 1088 reproduce the issue. Ideally, make the kernel's build configuration 1089 (.config) and the output from ``dmesg`` available somewhere on the net and 1090 link to it. Include or upload all other information that might be relevant, 1091 like the output/screenshot of an Oops or the output from ``lspci``. Once 1092 you wrote this main part, insert a normal length paragraph on top of it 1093 outlining the issue and the impact quickly. On top of this add one sentence 1094 that briefly describes the problem and gets people to read on. Now give the 1095 thing a descriptive title or subject that yet again is shorter. Then you're 1096 ready to send or file the report like the MAINTAINERS file told you, unless 1097 you are dealing with one of those 'issues of high priority': they need 1098 special care which is explained in 'Special handling for high priority 1099 issues' below.* 1100 1101Now that you have prepared everything it's time to write your report. How to do 1102that is partly explained by the three documents linked to in the preface above. 1103That's why this text will only mention a few of the essentials as well as 1104things specific to the Linux kernel. 1105 1106There is one thing that fits both categories: the most crucial parts of your 1107report are the title/subject, the first sentence, and the first paragraph. 1108Developers often get quite a lot of mail. They thus often just take a few 1109seconds to skim a mail before deciding to move on or look closer. Thus: the 1110better the top section of your report, the higher are the chances that someone 1111will look into it and help you. And that is why you should ignore them for now 1112and write the detailed report first. ;-) 1113 1114Things each report should mention 1115~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1116 1117Describe in detail how your issue happens with the fresh vanilla kernel you 1118installed. Try to include the step-by-step instructions you wrote and optimized 1119earlier that outline how you and ideally others can reproduce the issue; in 1120those rare cases where that's impossible try to describe what you did to 1121trigger it. 1122 1123Also include all the relevant information others might need to understand the 1124issue and its environment. What's actually needed depends a lot on the issue, 1125but there are some things you should include always: 1126 1127 * the output from ``cat /proc/version``, which contains the Linux kernel 1128 version number and the compiler it was built with. 1129 1130 * the Linux distribution the machine is running (``hostnamectl | grep 1131 "Operating System"``) 1132 1133 * the architecture of the CPU and the operating system (``uname -mi``) 1134 1135 * if you are dealing with a regression and performed a bisection, mention the 1136 subject and the commit-id of the change that is causing it. 1137 1138In a lot of cases it's also wise to make two more things available to those 1139that read your report: 1140 1141 * the configuration used for building your Linux kernel (the '.config' file) 1142 1143 * the kernel's messages that you get from ``dmesg`` written to a file. Make 1144 sure that it starts with a line like 'Linux version 5.8-1 1145 (foobar@example.com) (gcc (GCC) 10.2.1, GNU ld version 2.34) #1 SMP Mon Aug 1146 3 14:54:37 UTC 2020' If it's missing, then important messages from the first 1147 boot phase already got discarded. In this case instead consider using 1148 ``journalctl -b 0 -k``; alternatively you can also reboot, reproduce the 1149 issue and call ``dmesg`` right afterwards. 1150 1151These two files are big, that's why it's a bad idea to put them directly into 1152your report. If you are filing the issue in a bug tracker then attach them to 1153the ticket. If you report the issue by mail do not attach them, as that makes 1154the mail too large; instead do one of these things: 1155 1156 * Upload the files somewhere public (your website, a public file paste 1157 service, a ticket created just for this purpose on `bugzilla.kernel.org 1158 <https://bugzilla.kernel.org/>`_, ...) and include a link to them in your 1159 report. Ideally use something where the files stay available for years, as 1160 they could be useful to someone many years from now; this for example can 1161 happen if five or ten years from now a developer works on some code that was 1162 changed just to fix your issue. 1163 1164 * Put the files aside and mention you will send them later in individual 1165 replies to your own mail. Just remember to actually do that once the report 1166 went out. ;-) 1167 1168Things that might be wise to provide 1169~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1170 1171Depending on the issue you might need to add more background data. Here are a 1172few suggestions what often is good to provide: 1173 1174 * If you are dealing with a 'warning', an 'OOPS' or a 'panic' from the kernel, 1175 include it. If you can't copy'n'paste it, try to capture a netconsole trace 1176 or at least take a picture of the screen. 1177 1178 * If the issue might be related to your computer hardware, mention what kind 1179 of system you use. If you for example have problems with your graphics card, 1180 mention its manufacturer, the card's model, and what chip is uses. If it's a 1181 laptop mention its name, but try to make sure it's meaningful. 'Dell XPS 13' 1182 for example is not, because it might be the one from 2012; that one looks 1183 not that different from the one sold today, but apart from that the two have 1184 nothing in common. Hence, in such cases add the exact model number, which 1185 for example are '9380' or '7390' for XPS 13 models introduced during 2019. 1186 Names like 'Lenovo Thinkpad T590' are also somewhat ambiguous: there are 1187 variants of this laptop with and without a dedicated graphics chip, so try 1188 to find the exact model name or specify the main components. 1189 1190 * Mention the relevant software in use. If you have problems with loading 1191 modules, you want to mention the versions of kmod, systemd, and udev in use. 1192 If one of the DRM drivers misbehaves, you want to state the versions of 1193 libdrm and Mesa; also specify your Wayland compositor or the X-Server and 1194 its driver. If you have a filesystem issue, mention the version of 1195 corresponding filesystem utilities (e2fsprogs, btrfs-progs, xfsprogs, ...). 1196 1197 * Gather additional information from the kernel that might be of interest. The 1198 output from ``lspci -nn`` will for example help others to identify what 1199 hardware you use. If you have a problem with hardware you even might want to 1200 make the output from ``sudo lspci -vvv`` available, as that provides 1201 insights how the components were configured. For some issues it might be 1202 good to include the contents of files like ``/proc/cpuinfo``, 1203 ``/proc/ioports``, ``/proc/iomem``, ``/proc/modules``, or 1204 ``/proc/scsi/scsi``. Some subsystem also offer tools to collect relevant 1205 information. One such tool is ``alsa-info.sh`` `which the audio/sound 1206 subsystem developers provide <https://www.alsa-project.org/wiki/AlsaInfo>`_. 1207 1208Those examples should give your some ideas of what data might be wise to 1209attach, but you have to think yourself what will be helpful for others to know. 1210Don't worry too much about forgetting something, as developers will ask for 1211additional details they need. But making everything important available from 1212the start increases the chance someone will take a closer look. 1213 1214 1215The important part: the head of your report 1216~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1217 1218Now that you have the detailed part of the report prepared let's get to the 1219most important section: the first few sentences. Thus go to the top, add 1220something like 'The detailed description:' before the part you just wrote and 1221insert two newlines at the top. Now write one normal length paragraph that 1222describes the issue roughly. Leave out all boring details and focus on the 1223crucial parts readers need to know to understand what this is all about; if you 1224think this bug affects a lot of users, mention this to get people interested. 1225 1226Once you did that insert two more lines at the top and write a one sentence 1227summary that explains quickly what the report is about. After that you have to 1228get even more abstract and write an even shorter subject/title for the report. 1229 1230Now that you have written this part take some time to optimize it, as it is the 1231most important parts of your report: a lot of people will only read this before 1232they decide if reading the rest is time well spent. 1233 1234Now send or file the report like the :ref:`MAINTAINERS <maintainers>` file told 1235you, unless it's one of those 'issues of high priority' outlined earlier: in 1236that case please read the next subsection first before sending the report on 1237its way. 1238 1239Special handling for high priority issues 1240~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1241 1242Reports for high priority issues need special handling. 1243 1244**Severe issues**: make sure the subject or ticket title as well as the first 1245paragraph makes the severeness obvious. 1246 1247**Regressions**: make the report's subject start with '[REGRESSION]'. 1248 1249In case you performed a successful bisection, use the title of the change that 1250introduced the regression as the second part of your subject. Make the report 1251also mention the commit id of the culprit. In case of an unsuccessful bisection, 1252make your report mention the latest tested version that's working fine (say 5.7) 1253and the oldest where the issue occurs (say 5.8-rc1). 1254 1255When sending the report by mail, CC the Linux regressions mailing list 1256(regressions@lists.linux.dev). In case the report needs to be filed to some web 1257tracker, proceed to do so. Once filed, forward the report by mail to the 1258regressions list; CC the maintainer and the mailing list for the subsystem in 1259question. Make sure to inline the forwarded report, hence do not attach it. 1260Also add a short note at the top where you mention the URL to the ticket. 1261 1262When mailing or forwarding the report, in case of a successful bisection add the 1263author of the culprit to the recipients; also CC everyone in the signed-off-by 1264chain, which you find at the end of its commit message. 1265 1266**Security issues**: for these issues your will have to evaluate if a 1267short-term risk to other users would arise if details were publicly disclosed. 1268If that's not the case simply proceed with reporting the issue as described. 1269For issues that bear such a risk you will need to adjust the reporting process 1270slightly: 1271 1272 * If the MAINTAINERS file instructed you to report the issue by mail, do not 1273 CC any public mailing lists. 1274 1275 * If you were supposed to file the issue in a bug tracker make sure to mark 1276 the ticket as 'private' or 'security issue'. If the bug tracker does not 1277 offer a way to keep reports private, forget about it and send your report as 1278 a private mail to the maintainers instead. 1279 1280In both cases make sure to also mail your report to the addresses the 1281MAINTAINERS file lists in the section 'security contact'. Ideally directly CC 1282them when sending the report by mail. If you filed it in a bug tracker, forward 1283the report's text to these addresses; but on top of it put a small note where 1284you mention that you filed it with a link to the ticket. 1285 1286See 'Documentation/admin-guide/security-bugs.rst' for more information. 1287 1288 1289Duties after the report went out 1290-------------------------------- 1291 1292 *Wait for reactions and keep the thing rolling until you can accept the 1293 outcome in one way or the other. Thus react publicly and in a timely manner 1294 to any inquiries. Test proposed fixes. Do proactive testing: retest with at 1295 least every first release candidate (RC) of a new mainline version and 1296 report your results. Send friendly reminders if things stall. And try to 1297 help yourself, if you don't get any help or if it's unsatisfying.* 1298 1299If your report was good and you are really lucky then one of the developers 1300might immediately spot what's causing the issue; they then might write a patch 1301to fix it, test it, and send it straight for integration in mainline while 1302tagging it for later backport to stable and longterm kernels that need it. Then 1303all you need to do is reply with a 'Thank you very much' and switch to a version 1304with the fix once it gets released. 1305 1306But this ideal scenario rarely happens. That's why the job is only starting 1307once you got the report out. What you'll have to do depends on the situations, 1308but often it will be the things listed below. But before digging into the 1309details, here are a few important things you need to keep in mind for this part 1310of the process. 1311 1312 1313General advice for further interactions 1314~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1315 1316**Always reply in public**: When you filed the issue in a bug tracker, always 1317reply there and do not contact any of the developers privately about it. For 1318mailed reports always use the 'Reply-all' function when replying to any mails 1319you receive. That includes mails with any additional data you might want to add 1320to your report: go to your mail applications 'Sent' folder and use 'reply-all' 1321on your mail with the report. This approach will make sure the public mailing 1322list(s) and everyone else that gets involved over time stays in the loop; it 1323also keeps the mail thread intact, which among others is really important for 1324mailing lists to group all related mails together. 1325 1326There are just two situations where a comment in a bug tracker or a 'Reply-all' 1327is unsuitable: 1328 1329 * Someone tells you to send something privately. 1330 1331 * You were told to send something, but noticed it contains sensitive 1332 information that needs to be kept private. In that case it's okay to send it 1333 in private to the developer that asked for it. But note in the ticket or a 1334 mail that you did that, so everyone else knows you honored the request. 1335 1336**Do research before asking for clarifications or help**: In this part of the 1337process someone might tell you to do something that requires a skill you might 1338not have mastered yet. For example, you might be asked to use some test tools 1339you never have heard of yet; or you might be asked to apply a patch to the 1340Linux kernel sources to test if it helps. In some cases it will be fine sending 1341a reply asking for instructions how to do that. But before going that route try 1342to find the answer own your own by searching the internet; alternatively 1343consider asking in other places for advice. For example ask a friend or post 1344about it to a chatroom or forum you normally hang out. 1345 1346**Be patient**: If you are really lucky you might get a reply to your report 1347within a few hours. But most of the time it will take longer, as maintainers 1348are scattered around the globe and thus might be in a different time zone – one 1349where they already enjoy their night away from keyboard. 1350 1351In general, kernel developers will take one to five business days to respond to 1352reports. Sometimes it will take longer, as they might be busy with the merge 1353windows, other work, visiting developer conferences, or simply enjoying a long 1354summer holiday. 1355 1356The 'issues of high priority' (see above for an explanation) are an exception 1357here: maintainers should address them as soon as possible; that's why you 1358should wait a week at maximum (or just two days if it's something urgent) 1359before sending a friendly reminder. 1360 1361Sometimes the maintainer might not be responding in a timely manner; other 1362times there might be disagreements, for example if an issue qualifies as 1363regression or not. In such cases raise your concerns on the mailing list and 1364ask others for public or private replies how to move on. If that fails, it 1365might be appropriate to get a higher authority involved. In case of a WiFi 1366driver that would be the wireless maintainers; if there are no higher level 1367maintainers or all else fails, it might be one of those rare situations where 1368it's okay to get Linus Torvalds involved. 1369 1370**Proactive testing**: Every time the first pre-release (the 'rc1') of a new 1371mainline kernel version gets released, go and check if the issue is fixed there 1372or if anything of importance changed. Mention the outcome in the ticket or in a 1373mail you sent as reply to your report (make sure it has all those in the CC 1374that up to that point participated in the discussion). This will show your 1375commitment and that you are willing to help. It also tells developers if the 1376issue persists and makes sure they do not forget about it. A few other 1377occasional retests (for example with rc3, rc5 and the final) are also a good 1378idea, but only report your results if something relevant changed or if you are 1379writing something anyway. 1380 1381With all these general things off the table let's get into the details of how 1382to help to get issues resolved once they were reported. 1383 1384Inquires and testing request 1385~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1386 1387Here are your duties in case you got replies to your report: 1388 1389**Check who you deal with**: Most of the time it will be the maintainer or a 1390developer of the particular code area that will respond to your report. But as 1391issues are normally reported in public it could be anyone that's replying — 1392including people that want to help, but in the end might guide you totally off 1393track with their questions or requests. That rarely happens, but it's one of 1394many reasons why it's wise to quickly run an internet search to see who you're 1395interacting with. By doing this you also get aware if your report was heard by 1396the right people, as a reminder to the maintainer (see below) might be in order 1397later if discussion fades out without leading to a satisfying solution for the 1398issue. 1399 1400**Inquiries for data**: Often you will be asked to test something or provide 1401additional details. Try to provide the requested information soon, as you have 1402the attention of someone that might help and risk losing it the longer you 1403wait; that outcome is even likely if you do not provide the information within 1404a few business days. 1405 1406**Requests for testing**: When you are asked to test a diagnostic patch or a 1407possible fix, try to test it in timely manner, too. But do it properly and make 1408sure to not rush it: mixing things up can happen easily and can lead to a lot 1409of confusion for everyone involved. A common mistake for example is thinking a 1410proposed patch with a fix was applied, but in fact wasn't. Things like that 1411happen even to experienced testers occasionally, but they most of the time will 1412notice when the kernel with the fix behaves just as one without it. 1413 1414What to do when nothing of substance happens 1415~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1416 1417Some reports will not get any reaction from the responsible Linux kernel 1418developers; or a discussion around the issue evolved, but faded out with 1419nothing of substance coming out of it. 1420 1421In these cases wait two (better: three) weeks before sending a friendly 1422reminder: maybe the maintainer was just away from keyboard for a while when 1423your report arrived or had something more important to take care of. When 1424writing the reminder, kindly ask if anything else from your side is needed to 1425get the ball running somehow. If the report got out by mail, do that in the 1426first lines of a mail that is a reply to your initial mail (see above) which 1427includes a full quote of the original report below: that's on of those few 1428situations where such a 'TOFU' (Text Over, Fullquote Under) is the right 1429approach, as then all the recipients will have the details at hand immediately 1430in the proper order. 1431 1432After the reminder wait three more weeks for replies. If you still don't get a 1433proper reaction, you first should reconsider your approach. Did you maybe try 1434to reach out to the wrong people? Was the report maybe offensive or so 1435confusing that people decided to completely stay away from it? The best way to 1436rule out such factors: show the report to one or two people familiar with FLOSS 1437issue reporting and ask for their opinion. Also ask them for their advice how 1438to move forward. That might mean: prepare a better report and make those people 1439review it before you send it out. Such an approach is totally fine; just 1440mention that this is the second and improved report on the issue and include a 1441link to the first report. 1442 1443If the report was proper you can send a second reminder; in it ask for advice 1444why the report did not get any replies. A good moment for this second reminder 1445mail is shortly after the first pre-release (the 'rc1') of a new Linux kernel 1446version got published, as you should retest and provide a status update at that 1447point anyway (see above). 1448 1449If the second reminder again results in no reaction within a week, try to 1450contact a higher-level maintainer asking for advice: even busy maintainers by 1451then should at least have sent some kind of acknowledgment. 1452 1453Remember to prepare yourself for a disappointment: maintainers ideally should 1454react somehow to every issue report, but they are only obliged to fix those 1455'issues of high priority' outlined earlier. So don't be too devastating if you 1456get a reply along the lines of 'thanks for the report, I have more important 1457issues to deal with currently and won't have time to look into this for the 1458foreseeable future'. 1459 1460It's also possible that after some discussion in the bug tracker or on a list 1461nothing happens anymore and reminders don't help to motivate anyone to work out 1462a fix. Such situations can be devastating, but is within the cards when it 1463comes to Linux kernel development. This and several other reasons for not 1464getting help are explained in 'Why some issues won't get any reaction or remain 1465unfixed after being reported' near the end of this document. 1466 1467Don't get devastated if you don't find any help or if the issue in the end does 1468not get solved: the Linux kernel is FLOSS and thus you can still help yourself. 1469You for example could try to find others that are affected and team up with 1470them to get the issue resolved. Such a team could prepare a fresh report 1471together that mentions how many you are and why this is something that in your 1472option should get fixed. Maybe together you can also narrow down the root cause 1473or the change that introduced a regression, which often makes developing a fix 1474easier. And with a bit of luck there might be someone in the team that knows a 1475bit about programming and might be able to write a fix. 1476 1477 1478Reference for "Reporting regressions within a stable and longterm kernel line" 1479------------------------------------------------------------------------------ 1480 1481This subsection provides details for the steps you need to perform if you face 1482a regression within a stable and longterm kernel line. 1483 1484Make sure the particular version line still gets support 1485~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1486 1487 *Check if the kernel developers still maintain the Linux kernel version 1488 line you care about: go to the front page of kernel.org and make sure it 1489 mentions the latest release of the particular version line without an 1490 '[EOL]' tag.* 1491 1492Most kernel version lines only get supported for about three months, as 1493maintaining them longer is quite a lot of work. Hence, only one per year is 1494chosen and gets supported for at least two years (often six). That's why you 1495need to check if the kernel developers still support the version line you care 1496for. 1497 1498Note, if kernel.org lists two stable version lines on the front page, you 1499should consider switching to the newer one and forget about the older one: 1500support for it is likely to be abandoned soon. Then it will get a "end-of-life" 1501(EOL) stamp. Version lines that reached that point still get mentioned on the 1502kernel.org front page for a week or two, but are unsuitable for testing and 1503reporting. 1504 1505Search stable mailing list 1506~~~~~~~~~~~~~~~~~~~~~~~~~~ 1507 1508 *Check the archives of the Linux stable mailing list for existing reports.* 1509 1510Maybe the issue you face is already known and was fixed or is about to. Hence, 1511`search the archives of the Linux stable mailing list 1512<https://lore.kernel.org/stable/>`_ for reports about an issue like yours. If 1513you find any matches, consider joining the discussion, unless the fix is 1514already finished and scheduled to get applied soon. 1515 1516Reproduce issue with the newest release 1517~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1518 1519 *Install the latest release from the particular version line as a vanilla 1520 kernel. Ensure this kernel is not tainted and still shows the problem, as 1521 the issue might have already been fixed there. If you first noticed the 1522 problem with a vendor kernel, check a vanilla build of the last version 1523 known to work performs fine as well.* 1524 1525Before investing any more time in this process you want to check if the issue 1526was already fixed in the latest release of version line you're interested in. 1527This kernel needs to be vanilla and shouldn't be tainted before the issue 1528happens, as detailed outlined already above in the section "Install a fresh 1529kernel for testing". 1530 1531Did you first notice the regression with a vendor kernel? Then changes the 1532vendor applied might be interfering. You need to rule that out by performing 1533a recheck. Say something broke when you updated from 5.10.4-vendor.42 to 15345.10.5-vendor.43. Then after testing the latest 5.10 release as outlined in 1535the previous paragraph check if a vanilla build of Linux 5.10.4 works fine as 1536well. If things are broken there, the issue does not qualify as upstream 1537regression and you need switch back to the main step-by-step guide to report 1538the issue. 1539 1540Report the regression 1541~~~~~~~~~~~~~~~~~~~~~ 1542 1543 *Send a short problem report to the Linux stable mailing list 1544 (stable@vger.kernel.org) and CC the Linux regressions mailing list 1545 (regressions@lists.linux.dev); if you suspect the cause in a particular 1546 subsystem, CC its maintainer and its mailing list. Roughly describe the 1547 issue and ideally explain how to reproduce it. Mention the first version 1548 that shows the problem and the last version that's working fine. Then 1549 wait for further instructions.* 1550 1551When reporting a regression that happens within a stable or longterm kernel 1552line (say when updating from 5.10.4 to 5.10.5) a brief report is enough for 1553the start to get the issue reported quickly. Hence a rough description to the 1554stable and regressions mailing list is all it takes; but in case you suspect 1555the cause in a particular subsystem, CC its maintainers and its mailing list 1556as well, because that will speed things up. 1557 1558And note, it helps developers a great deal if you can specify the exact version 1559that introduced the problem. Hence if possible within a reasonable time frame, 1560try to find that version using vanilla kernels. Lets assume something broke when 1561your distributor released a update from Linux kernel 5.10.5 to 5.10.8. Then as 1562instructed above go and check the latest kernel from that version line, say 15635.10.9. If it shows the problem, try a vanilla 5.10.5 to ensure that no patches 1564the distributor applied interfere. If the issue doesn't manifest itself there, 1565try 5.10.7 and then (depending on the outcome) 5.10.8 or 5.10.6 to find the 1566first version where things broke. Mention it in the report and state that 5.10.9 1567is still broken. 1568 1569What the previous paragraph outlines is basically a rough manual 'bisection'. 1570Once your report is out your might get asked to do a proper one, as it allows to 1571pinpoint the exact change that causes the issue (which then can easily get 1572reverted to fix the issue quickly). Hence consider to do a proper bisection 1573right away if time permits. See the section 'Special care for regressions' and 1574the document 'Documentation/admin-guide/bug-bisect.rst' for details how to 1575perform one. In case of a successful bisection add the author of the culprit to 1576the recipients; also CC everyone in the signed-off-by chain, which you find at 1577the end of its commit message. 1578 1579 1580Reference for "Reporting issues only occurring in older kernel version lines" 1581----------------------------------------------------------------------------- 1582 1583This section provides details for the steps you need to take if you could not 1584reproduce your issue with a mainline kernel, but want to see it fixed in older 1585version lines (aka stable and longterm kernels). 1586 1587Some fixes are too complex 1588~~~~~~~~~~~~~~~~~~~~~~~~~~ 1589 1590 *Prepare yourself for the possibility that going through the next few steps 1591 might not get the issue solved in older releases: the fix might be too big 1592 or risky to get backported there.* 1593 1594Even small and seemingly obvious code-changes sometimes introduce new and 1595totally unexpected problems. The maintainers of the stable and longterm kernels 1596are very aware of that and thus only apply changes to these kernels that are 1597within rules outlined in 'Documentation/process/stable-kernel-rules.rst'. 1598 1599Complex or risky changes for example do not qualify and thus only get applied 1600to mainline. Other fixes are easy to get backported to the newest stable and 1601longterm kernels, but too risky to integrate into older ones. So be aware the 1602fix you are hoping for might be one of those that won't be backported to the 1603version line your care about. In that case you'll have no other choice then to 1604live with the issue or switch to a newer Linux version, unless you want to 1605patch the fix into your kernels yourself. 1606 1607Common preparations 1608~~~~~~~~~~~~~~~~~~~ 1609 1610 *Perform the first three steps in the section "Reporting issues only 1611 occurring in older kernel version lines" above.* 1612 1613You need to carry out a few steps already described in another section of this 1614guide. Those steps will let you: 1615 1616 * Check if the kernel developers still maintain the Linux kernel version line 1617 you care about. 1618 1619 * Search the Linux stable mailing list for exiting reports. 1620 1621 * Check with the latest release. 1622 1623 1624Check code history and search for existing discussions 1625~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1626 1627 *Search the Linux kernel version control system for the change that fixed 1628 the issue in mainline, as its commit message might tell you if the fix is 1629 scheduled for backporting already. If you don't find anything that way, 1630 search the appropriate mailing lists for posts that discuss such an issue 1631 or peer-review possible fixes; then check the discussions if the fix was 1632 deemed unsuitable for backporting. If backporting was not considered at 1633 all, join the newest discussion, asking if it's in the cards.* 1634 1635In a lot of cases the issue you deal with will have happened with mainline, but 1636got fixed there. The commit that fixed it would need to get backported as well 1637to get the issue solved. That's why you want to search for it or any 1638discussions abound it. 1639 1640 * First try to find the fix in the Git repository that holds the Linux kernel 1641 sources. You can do this with the web interfaces `on kernel.org 1642 <https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/>`_ 1643 or its mirror `on GitHub <https://github.com/torvalds/linux>`_; if you have 1644 a local clone you alternatively can search on the command line with ``git 1645 log --grep=<pattern>``. 1646 1647 If you find the fix, look if the commit message near the end contains a 1648 'stable tag' that looks like this: 1649 1650 Cc: <stable@vger.kernel.org> # 5.4+ 1651 1652 If that's case the developer marked the fix safe for backporting to version 1653 line 5.4 and later. Most of the time it's getting applied there within two 1654 weeks, but sometimes it takes a bit longer. 1655 1656 * If the commit doesn't tell you anything or if you can't find the fix, look 1657 again for discussions about the issue. Search the net with your favorite 1658 internet search engine as well as the archives for the `Linux kernel 1659 developers mailing list <https://lore.kernel.org/lkml/>`_. Also read the 1660 section `Locate kernel area that causes the issue` above and follow the 1661 instructions to find the subsystem in question: its bug tracker or mailing 1662 list archive might have the answer you are looking for. 1663 1664 * If you see a proposed fix, search for it in the version control system as 1665 outlined above, as the commit might tell you if a backport can be expected. 1666 1667 * Check the discussions for any indicators the fix might be too risky to get 1668 backported to the version line you care about. If that's the case you have 1669 to live with the issue or switch to the kernel version line where the fix 1670 got applied. 1671 1672 * If the fix doesn't contain a stable tag and backporting was not discussed, 1673 join the discussion: mention the version where you face the issue and that 1674 you would like to see it fixed, if suitable. 1675 1676 1677Ask for advice 1678~~~~~~~~~~~~~~ 1679 1680 *One of the former steps should lead to a solution. If that doesn't work 1681 out, ask the maintainers for the subsystem that seems to be causing the 1682 issue for advice; CC the mailing list for the particular subsystem as well 1683 as the stable mailing list.* 1684 1685If the previous three steps didn't get you closer to a solution there is only 1686one option left: ask for advice. Do that in a mail you sent to the maintainers 1687for the subsystem where the issue seems to have its roots; CC the mailing list 1688for the subsystem as well as the stable mailing list (stable@vger.kernel.org). 1689 1690 1691Why some issues won't get any reaction or remain unfixed after being reported 1692============================================================================= 1693 1694When reporting a problem to the Linux developers, be aware only 'issues of high 1695priority' (regressions, security issues, severe problems) are definitely going 1696to get resolved. The maintainers or if all else fails Linus Torvalds himself 1697will make sure of that. They and the other kernel developers will fix a lot of 1698other issues as well. But be aware that sometimes they can't or won't help; and 1699sometimes there isn't even anyone to send a report to. 1700 1701This is best explained with kernel developers that contribute to the Linux 1702kernel in their spare time. Quite a few of the drivers in the kernel were 1703written by such programmers, often because they simply wanted to make their 1704hardware usable on their favorite operating system. 1705 1706These programmers most of the time will happily fix problems other people 1707report. But nobody can force them to do, as they are contributing voluntarily. 1708 1709Then there are situations where such developers really want to fix an issue, 1710but can't: sometimes they lack hardware programming documentation to do so. 1711This often happens when the publicly available docs are superficial or the 1712driver was written with the help of reverse engineering. 1713 1714Sooner or later spare time developers will also stop caring for the driver. 1715Maybe their test hardware broke, got replaced by something more fancy, or is so 1716old that it's something you don't find much outside of computer museums 1717anymore. Sometimes developer stops caring for their code and Linux at all, as 1718something different in their life became way more important. In some cases 1719nobody is willing to take over the job as maintainer – and nobody can be forced 1720to, as contributing to the Linux kernel is done on a voluntary basis. Abandoned 1721drivers nevertheless remain in the kernel: they are still useful for people and 1722removing would be a regression. 1723 1724The situation is not that different with developers that are paid for their 1725work on the Linux kernel. Those contribute most changes these days. But their 1726employers sooner or later also stop caring for their code or make its 1727programmer focus on other things. Hardware vendors for example earn their money 1728mainly by selling new hardware; quite a few of them hence are not investing 1729much time and energy in maintaining a Linux kernel driver for something they 1730stopped selling years ago. Enterprise Linux distributors often care for a 1731longer time period, but in new versions often leave support for old and rare 1732hardware aside to limit the scope. Often spare time contributors take over once 1733a company orphans some code, but as mentioned above: sooner or later they will 1734leave the code behind, too. 1735 1736Priorities are another reason why some issues are not fixed, as maintainers 1737quite often are forced to set those, as time to work on Linux is limited. 1738That's true for spare time or the time employers grant their developers to 1739spend on maintenance work on the upstream kernel. Sometimes maintainers also 1740get overwhelmed with reports, even if a driver is working nearly perfectly. To 1741not get completely stuck, the programmer thus might have no other choice than 1742to prioritize issue reports and reject some of them. 1743 1744But don't worry too much about all of this, a lot of drivers have active 1745maintainers who are quite interested in fixing as many issues as possible. 1746 1747 1748Closing words 1749============= 1750 1751Compared with other Free/Libre & Open Source Software it's hard to report 1752issues to the Linux kernel developers: the length and complexity of this 1753document and the implications between the lines illustrate that. But that's how 1754it is for now. The main author of this text hopes documenting the state of the 1755art will lay some groundwork to improve the situation over time. 1756 1757 1758.. 1759 This text is maintained by Thorsten Leemhuis <linux@leemhuis.info>. If you 1760 spot a typo or small mistake, feel free to let him know directly and he'll 1761 fix it. You are free to do the same in a mostly informal way if you want 1762 to contribute changes to the text, but for copyright reasons please CC 1763 linux-doc@vger.kernel.org and "sign-off" your contribution as 1764 Documentation/process/submitting-patches.rst outlines in the section "Sign 1765 your work - the Developer's Certificate of Origin". 1766