Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 1 | Embargoed hardware issues |
| 2 | ========================= |
| 3 | |
| 4 | Scope |
| 5 | ----- |
| 6 | |
| 7 | Hardware issues which result in security problems are a different category |
Kees Cook | f56f791 | 2019-09-04 09:24:49 -0700 | [diff] [blame] | 8 | of security bugs than pure software bugs which only affect the Linux |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 9 | kernel. |
| 10 | |
| 11 | Hardware issues like Meltdown, Spectre, L1TF etc. must be treated |
| 12 | differently because they usually affect all Operating Systems ("OS") and |
| 13 | therefore need coordination across different OS vendors, distributions, |
| 14 | hardware vendors and other parties. For some of the issues, software |
| 15 | mitigations can depend on microcode or firmware updates, which need further |
| 16 | coordination. |
| 17 | |
| 18 | .. _Contact: |
| 19 | |
| 20 | Contact |
| 21 | ------- |
| 22 | |
| 23 | The Linux kernel hardware security team is separate from the regular Linux |
| 24 | kernel security team. |
| 25 | |
| 26 | The team only handles the coordination of embargoed hardware security |
| 27 | issues. Reports of pure software security bugs in the Linux kernel are not |
| 28 | handled by this team and the reporter will be guided to contact the regular |
| 29 | Linux kernel security team (:ref:`Documentation/admin-guide/ |
| 30 | <securitybugs>`) instead. |
| 31 | |
| 32 | The team can be contacted by email at <hardware-security@kernel.org>. This |
| 33 | is a private list of security officers who will help you to coordinate an |
| 34 | issue according to our documented process. |
| 35 | |
| 36 | The list is encrypted and email to the list can be sent by either PGP or |
| 37 | S/MIME encrypted and must be signed with the reporter's PGP key or S/MIME |
| 38 | certificate. The list's PGP key and S/MIME certificate are available from |
| 39 | https://www.kernel.org/.... |
| 40 | |
| 41 | While hardware security issues are often handled by the affected hardware |
| 42 | vendor, we welcome contact from researchers or individuals who have |
| 43 | identified a potential hardware flaw. |
| 44 | |
| 45 | Hardware security officers |
| 46 | ^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| 47 | |
| 48 | The current team of hardware security officers: |
| 49 | |
| 50 | - Linus Torvalds (Linux Foundation Fellow) |
| 51 | - Greg Kroah-Hartman (Linux Foundation Fellow) |
| 52 | - Thomas Gleixner (Linux Foundation Fellow) |
| 53 | |
| 54 | Operation of mailing-lists |
| 55 | ^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| 56 | |
| 57 | The encrypted mailing-lists which are used in our process are hosted on |
| 58 | Linux Foundation's IT infrastructure. By providing this service Linux |
| 59 | Foundation's director of IT Infrastructure security technically has the |
| 60 | ability to access the embargoed information, but is obliged to |
| 61 | confidentiality by his employment contract. Linux Foundation's director of |
| 62 | IT Infrastructure security is also responsible for the kernel.org |
| 63 | infrastructure. |
| 64 | |
| 65 | The Linux Foundation's current director of IT Infrastructure security is |
| 66 | Konstantin Ryabitsev. |
| 67 | |
| 68 | |
| 69 | Non-disclosure agreements |
| 70 | ------------------------- |
| 71 | |
| 72 | The Linux kernel hardware security team is not a formal body and therefore |
| 73 | unable to enter into any non-disclosure agreements. The kernel community |
| 74 | is aware of the sensitive nature of such issues and offers a Memorandum of |
| 75 | Understanding instead. |
| 76 | |
| 77 | |
| 78 | Memorandum of Understanding |
| 79 | --------------------------- |
| 80 | |
| 81 | The Linux kernel community has a deep understanding of the requirement to |
| 82 | keep hardware security issues under embargo for coordination between |
| 83 | different OS vendors, distributors, hardware vendors and other parties. |
| 84 | |
| 85 | The Linux kernel community has successfully handled hardware security |
| 86 | issues in the past and has the necessary mechanisms in place to allow |
| 87 | community compliant development under embargo restrictions. |
| 88 | |
| 89 | The Linux kernel community has a dedicated hardware security team for |
| 90 | initial contact, which oversees the process of handling such issues under |
| 91 | embargo rules. |
| 92 | |
| 93 | The hardware security team identifies the developers (domain experts) who |
| 94 | will form the initial response team for a particular issue. The initial |
| 95 | response team can bring in further developers (domain experts) to address |
| 96 | the issue in the best technical way. |
| 97 | |
| 98 | All involved developers pledge to adhere to the embargo rules and to keep |
| 99 | the received information confidential. Violation of the pledge will lead to |
| 100 | immediate exclusion from the current issue and removal from all related |
| 101 | mailing-lists. In addition, the hardware security team will also exclude |
| 102 | the offender from future issues. The impact of this consequence is a highly |
| 103 | effective deterrent in our community. In case a violation happens the |
| 104 | hardware security team will inform the involved parties immediately. If you |
| 105 | or anyone becomes aware of a potential violation, please report it |
| 106 | immediately to the Hardware security officers. |
| 107 | |
| 108 | |
| 109 | Process |
| 110 | ^^^^^^^ |
| 111 | |
| 112 | Due to the globally distributed nature of Linux kernel development, |
| 113 | face-to-face meetings are almost impossible to address hardware security |
| 114 | issues. Phone conferences are hard to coordinate due to time zones and |
| 115 | other factors and should be only used when absolutely necessary. Encrypted |
| 116 | email has been proven to be the most effective and secure communication |
| 117 | method for these types of issues. |
| 118 | |
| 119 | Start of Disclosure |
| 120 | """"""""""""""""""" |
| 121 | |
| 122 | Disclosure starts by contacting the Linux kernel hardware security team by |
| 123 | email. This initial contact should contain a description of the problem and |
| 124 | a list of any known affected hardware. If your organization builds or |
| 125 | distributes the affected hardware, we encourage you to also consider what |
| 126 | other hardware could be affected. |
| 127 | |
| 128 | The hardware security team will provide an incident-specific encrypted |
| 129 | mailing-list which will be used for initial discussion with the reporter, |
| 130 | further disclosure and coordination. |
| 131 | |
| 132 | The hardware security team will provide the disclosing party a list of |
| 133 | developers (domain experts) who should be informed initially about the |
| 134 | issue after confirming with the developers that they will adhere to this |
| 135 | Memorandum of Understanding and the documented process. These developers |
| 136 | form the initial response team and will be responsible for handling the |
| 137 | issue after initial contact. The hardware security team is supporting the |
| 138 | response team, but is not necessarily involved in the mitigation |
| 139 | development process. |
| 140 | |
| 141 | While individual developers might be covered by a non-disclosure agreement |
| 142 | via their employer, they cannot enter individual non-disclosure agreements |
| 143 | in their role as Linux kernel developers. They will, however, agree to |
| 144 | adhere to this documented process and the Memorandum of Understanding. |
| 145 | |
Thomas Gleixner | dc925a3 | 2019-09-25 10:29:49 +0200 | [diff] [blame] | 146 | The disclosing party should provide a list of contacts for all other |
| 147 | entities who have already been, or should be, informed about the issue. |
| 148 | This serves several purposes: |
| 149 | |
| 150 | - The list of disclosed entities allows communication accross the |
| 151 | industry, e.g. other OS vendors, HW vendors, etc. |
| 152 | |
| 153 | - The disclosed entities can be contacted to name experts who should |
| 154 | participate in the mitigation development. |
| 155 | |
| 156 | - If an expert which is required to handle an issue is employed by an |
| 157 | listed entity or member of an listed entity, then the response teams can |
| 158 | request the disclosure of that expert from that entity. This ensures |
| 159 | that the expert is also part of the entity's response team. |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 160 | |
| 161 | Disclosure |
| 162 | """""""""" |
| 163 | |
| 164 | The disclosing party provides detailed information to the initial response |
| 165 | team via the specific encrypted mailing-list. |
| 166 | |
| 167 | From our experience the technical documentation of these issues is usually |
| 168 | a sufficient starting point and further technical clarification is best |
| 169 | done via email. |
| 170 | |
| 171 | Mitigation development |
| 172 | """""""""""""""""""""" |
| 173 | |
| 174 | The initial response team sets up an encrypted mailing-list or repurposes |
Thomas Gleixner | dc925a3 | 2019-09-25 10:29:49 +0200 | [diff] [blame] | 175 | an existing one if appropriate. |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 176 | |
| 177 | Using a mailing-list is close to the normal Linux development process and |
| 178 | has been successfully used in developing mitigations for various hardware |
| 179 | security issues in the past. |
| 180 | |
| 181 | The mailing-list operates in the same way as normal Linux development. |
| 182 | Patches are posted, discussed and reviewed and if agreed on applied to a |
| 183 | non-public git repository which is only accessible to the participating |
| 184 | developers via a secure connection. The repository contains the main |
| 185 | development branch against the mainline kernel and backport branches for |
| 186 | stable kernel versions as necessary. |
| 187 | |
| 188 | The initial response team will identify further experts from the Linux |
Thomas Gleixner | dc925a3 | 2019-09-25 10:29:49 +0200 | [diff] [blame] | 189 | kernel developer community as needed. Bringing in experts can happen at any |
| 190 | time of the development process and needs to be handled in a timely manner. |
| 191 | |
| 192 | If an expert is employed by or member of an entity on the disclosure list |
| 193 | provided by the disclosing party, then participation will be requested from |
| 194 | the relevant entity. |
| 195 | |
| 196 | If not, then the disclosing party will be informed about the experts |
| 197 | participation. The experts are covered by the Memorandum of Understanding |
| 198 | and the disclosing party is requested to acknowledge the participation. In |
| 199 | case that the disclosing party has a compelling reason to object, then this |
| 200 | objection has to be raised within five work days and resolved with the |
| 201 | incident team immediately. If the disclosing party does not react within |
| 202 | five work days this is taken as silent acknowledgement. |
| 203 | |
| 204 | After acknowledgement or resolution of an objection the expert is disclosed |
| 205 | by the incident team and brought into the development process. |
| 206 | |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 207 | |
| 208 | Coordinated release |
| 209 | """"""""""""""""""" |
| 210 | |
| 211 | The involved parties will negotiate the date and time where the embargo |
| 212 | ends. At that point the prepared mitigations are integrated into the |
| 213 | relevant kernel trees and published. |
| 214 | |
| 215 | While we understand that hardware security issues need coordinated embargo |
| 216 | time, the embargo time should be constrained to the minimum time which is |
| 217 | required for all involved parties to develop, test and prepare the |
| 218 | mitigations. Extending embargo time artificially to meet conference talk |
| 219 | dates or other non-technical reasons is creating more work and burden for |
| 220 | the involved developers and response teams as the patches need to be kept |
| 221 | up to date in order to follow the ongoing upstream kernel development, |
| 222 | which might create conflicting changes. |
| 223 | |
| 224 | CVE assignment |
| 225 | """""""""""""" |
| 226 | |
| 227 | Neither the hardware security team nor the initial response team assign |
| 228 | CVEs, nor are CVEs required for the development process. If CVEs are |
| 229 | provided by the disclosing party they can be used for documentation |
| 230 | purposes. |
| 231 | |
| 232 | Process ambassadors |
| 233 | ------------------- |
| 234 | |
| 235 | For assistance with this process we have established ambassadors in various |
| 236 | organizations, who can answer questions about or provide guidance on the |
| 237 | reporting process and further handling. Ambassadors are not involved in the |
| 238 | disclosure of a particular issue, unless requested by a response team or by |
| 239 | an involved disclosed party. The current ambassadors list: |
| 240 | |
| 241 | ============= ======================================================== |
| 242 | ARM |
| 243 | AMD |
| 244 | IBM |
Tony Luck | 38c7a30 | 2019-09-10 10:26:46 -0700 | [diff] [blame] | 245 | Intel Tony Luck <tony.luck@intel.com> |
Trilok Soni | a8e0aba | 2019-09-06 12:01:57 -0700 | [diff] [blame] | 246 | Qualcomm Trilok Soni <tsoni@codeaurora.org> |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 247 | |
Sasha Levin | 1f49316 | 2019-09-06 05:58:52 -0400 | [diff] [blame] | 248 | Microsoft Sasha Levin <sashal@kernel.org> |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 249 | VMware |
Andrew Cooper | 02e740a | 2019-09-04 19:17:02 +0100 | [diff] [blame] | 250 | Xen Andrew Cooper <andrew.cooper3@citrix.com> |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 251 | |
| 252 | Canonical Tyler Hicks <tyhicks@canonical.com> |
| 253 | Debian Ben Hutchings <ben@decadent.org.uk> |
| 254 | Oracle Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> |
| 255 | Red Hat Josh Poimboeuf <jpoimboe@redhat.com> |
| 256 | SUSE Jiri Kosina <jkosina@suse.cz> |
| 257 | |
| 258 | Amazon |
Kees Cook | f56f791 | 2019-09-04 09:24:49 -0700 | [diff] [blame] | 259 | Google Kees Cook <keescook@chromium.org> |
| 260 | ============= ======================================================== |
Thomas Gleixner | ddaedbb | 2019-08-15 23:25:05 +0200 | [diff] [blame] | 261 | |
| 262 | If you want your organization to be added to the ambassadors list, please |
| 263 | contact the hardware security team. The nominated ambassador has to |
| 264 | understand and support our process fully and is ideally well connected in |
| 265 | the Linux kernel community. |
| 266 | |
| 267 | Encrypted mailing-lists |
| 268 | ----------------------- |
| 269 | |
| 270 | We use encrypted mailing-lists for communication. The operating principle |
| 271 | of these lists is that email sent to the list is encrypted either with the |
| 272 | list's PGP key or with the list's S/MIME certificate. The mailing-list |
| 273 | software decrypts the email and re-encrypts it individually for each |
| 274 | subscriber with the subscriber's PGP key or S/MIME certificate. Details |
| 275 | about the mailing-list software and the setup which is used to ensure the |
| 276 | security of the lists and protection of the data can be found here: |
| 277 | https://www.kernel.org/.... |
| 278 | |
| 279 | List keys |
| 280 | ^^^^^^^^^ |
| 281 | |
| 282 | For initial contact see :ref:`Contact`. For incident specific mailing-lists |
| 283 | the key and S/MIME certificate are conveyed to the subscribers by email |
| 284 | sent from the specific list. |
| 285 | |
| 286 | Subscription to incident specific lists |
| 287 | ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ |
| 288 | |
| 289 | Subscription is handled by the response teams. Disclosed parties who want |
| 290 | to participate in the communication send a list of potential subscribers to |
| 291 | the response team so the response team can validate subscription requests. |
| 292 | |
| 293 | Each subscriber needs to send a subscription request to the response team |
| 294 | by email. The email must be signed with the subscriber's PGP key or S/MIME |
| 295 | certificate. If a PGP key is used, it must be available from a public key |
| 296 | server and is ideally connected to the Linux kernel's PGP web of trust. See |
| 297 | also: https://www.kernel.org/signature.html. |
| 298 | |
| 299 | The response team verifies that the subscriber request is valid and adds |
| 300 | the subscriber to the list. After subscription the subscriber will receive |
| 301 | email from the mailing-list which is signed either with the list's PGP key |
| 302 | or the list's S/MIME certificate. The subscriber's email client can extract |
| 303 | the PGP key or the S/MIME certificate from the signature so the subscriber |
| 304 | can send encrypted email to the list. |
| 305 | |