summaryrefslogtreecommitdiff
path: root/drivers/interconnect
AgeCommit message (Collapse)Author
4 daysMerge branch 'icc-fixes' into icc-nextGeorgi Djakov
* icc-fixes Revert "interconnect: qcom: x1e80100: enable QoS configuration" Signed-off-by: Georgi Djakov <djakov@kernel.org>
4 daysMerge branch 'icc-misc' into icc-nextGeorgi Djakov
* icc-misc interconnect: qcom: Constify struct qcom_icc_qosbox interconnect: fix typo "commited" in comments interconnect: Add kernel-doc for devm_of_icc_get(), icc_enable() and icc_disable() interconnect: exynos: Remove the PM QoS request on unbind interconnect: qcom: implement get_bw with rpmh_read interconnect: qcom: Skip get_bw on QoS programming providers interconnect: debugfs: replace writable string helper interconnect: qcom: fix endian annotations of BCM aux data interconnect: mediatek: fix Makefile typo for mt8196 Signed-off-by: Georgi Djakov <djakov@kernel.org>
4 daysinterconnect: mediatek: fix Makefile typo for mt8196Breno Rodrigues Alves
Correct a copy-paste typo in the MediaTek interconnect Makefile that mapped mt8196.o to CONFIG_INTERCONNECT_MTK_MT8195 instead of MT8196. Assisted-by: OpenCode AI Signed-off-by: Breno Rodrigues Alves <breno3011alves@gmail.com> Link: https://patch.msgid.link/20260905140113.21638-1-breno3011alves@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
4 daysinterconnect: qcom: fix endian annotations of BCM aux dataRosen Penev
struct qcom_icc_bcm::aux_data keeps a copy of the struct bcm_db read from the command db, whose unit and width fields are annotated as __le32/__le16 to describe the little-endian on-disk format. Using those restricted types directly in bandwidth calculations makes sparse complain about endianness. Keep aux_data typed as struct bcm_db, copied verbatim from the command db buffer, and convert the fields with le32_to_cpu()/le16_to_cpu() at the places where unit and width are used. No functional change. Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202609221350.3Y8MOce9-lkp@intel.com/ Assisted-by: LLM Signed-off-by: Rosen Penev <rosenp@gmail.com> Link: https://patch.msgid.link/20260923185059.25196-1-rosenp@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
4 daysinterconnect: debugfs: replace writable string helperYichong Chen
debugfs_create_str() is being made read-only because its generic write path is hard to make safe without adding more locking to the helper. Convert the interconnect debugfs client src_node and dst_node entries to local file operations before removing writable string support from debugfs_create_str(). Protect the string replacement and path lookup with the existing debugfs_lock. The old code duplicated the strings under rcu_read_lock(), so it had to use GFP_ATOMIC. The local file operations protect src_node and dst_node with debugfs_lock instead, so the allocation can use GFP_KERNEL. Signed-off-by: Yichong Chen <chenyichong@uniontech.com> Link: https://patch.msgid.link/20260807100053.1089834-2-chenyichong@uniontech.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
10 daysinterconnect: qcom: Skip get_bw on QoS programming providersShawn Guo
Programming the static QoS configuration in qcom_icc_rpmh_probe() touches NoC registers directly, and nothing in that path clocks or powers the NoC first. It has been working only because the interconnect core used to vote every node to INT_MAX while adding it, and that vote stays in hardware until sync_state. Reading the boot-time AMC votes back from RPMh removed that implicit guarantee: a NoC whose BCMs come out of the bootloader unvoted is now left unclocked, and the first QoS register access aborts. On Qualcomm Nord SoC, the MMSS NoC dies while programming qnm_camnoc_hf, taking the board down before rootfs: Internal error: synchronous external abort: 0000000096001610 regmap_mmio_read32le+0xc/0xa4 regmap_update_bits_base+0x64/0x98 qcom_icc_rpmh_probe+0x44c/0x540 Only install the get_bw() callback for providers that do not program QoS, i.e. those without a regmap config. Providers that do program QoS keep the previous INT_MAX initial votes, so their register accesses stay safe, and the rest keep the benefit of honouring the boot-time votes. Reported-by: Bartosz Golaszewski <brgl@kernel.org> Fixes: 11a44c6087c6 ("interconnect: qcom: implement get_bw with rpmh_read") Assisted-by: LLM Signed-off-by: Shawn Guo <shengchao.guo@oss.qualcomm.com> Reviewed-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com> Tested-by: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com> Link: https://patch.msgid.link/20260922020623.398913-1-shengchao.guo@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysMerge branch 'icc-rpm' into icc-nextGeorgi Djakov
This series brings multi-port QoS programming and squashes some minor existing issues. * icc-rpm interconnect: qcom: Move bcm_div() to common code interconnect: qcom: icc-rpm: Don't skip small votes interconnect: qcom: icc-rpm: Handle icc_link_create() failures interconnect: qcom: icc-rpm: Program SMD RPM auxiliary QoS ports interconnect: qcom: msm8996: Configure auxiliary Venus QoS port interconnect: qcom: msm8976: Configure auxiliary GPU QoS port interconnect: qcom: use 64-bit SMD RPM bus votes Link: https://patch.msgid.link/20260909-topic-icc_rpm_smd_qol-v1-0-6dc5b5e97626@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysinterconnect: qcom: use 64-bit SMD RPM bus votesKonrad Dybcio
SMD RPM bus bandwidth requests carry a 64-bit bandwidth value. The interconnect proxy declared the KVP payload as u32, truncating votes above 4 GiB/s and advertising the wrong payload size to RPM. Use a little-endian 64-bit payload and accept a u64 value in the proxy API. This has always been the right behavior, even on msm-3.10 and 32-bit SoCs. Fixes: be06f8e7425d ("interconnect: qcom: Add interconnect RPM over SMD driver") Assisted-by: LLM Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Reviewed-by: Gopikrishna Garmidi <gopikrishna.garmidi@oss.qualcomm.com> Link: https://patch.msgid.link/20260904-topic-icc_rpm_64b_req-v1-1-37e24f1991a3@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysinterconnect: qcom: msm8976: Configure auxiliary GPU QoS portKonrad Dybcio
Enable the secondary QoS port used by the MSM8976 GPU node. Assisted-by: LLM Fixes: 6e3ea1f36b10 ("interconnect: qcom: Add MSM8976 interconnect provider driver") Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Link: https://patch.msgid.link/20260909-topic-icc_rpm_smd_qol-v1-6-6dc5b5e97626@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysinterconnect: qcom: msm8996: Configure auxiliary Venus QoS portKonrad Dybcio
Enable the secondary QoS port used by the MSM8996 Venus node. Assisted-by: LLM Fixes: 7add937f5222 ("interconnect: qcom: Add MSM8996 interconnect provider driver") Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Link: https://patch.msgid.link/20260909-topic-icc_rpm_smd_qol-v1-5-6dc5b5e97626@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysinterconnect: qcom: icc-rpm: Program SMD RPM auxiliary QoS portsKonrad Dybcio
Some SMD RPM interconnect nodes own two QoS ports. Currently, the code only takes care of one, leaving some endpoints partially unconfigured. Keep qos_port as the primary port and describe the optional second port with aux_qos_port. Assisted-by: LLM Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Link: https://patch.msgid.link/20260909-topic-icc_rpm_smd_qol-v1-4-6dc5b5e97626@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysinterconnect: qcom: icc-rpm: Handle icc_link_create() failuresKonrad Dybcio
icc_link_create() can return errors. Don't ignore them. Assisted-by: LLM Fixes: 30c8fa3ec61a ("interconnect: qcom: Add MSM8916 interconnect provider driver") Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Link: https://patch.msgid.link/20260909-topic-icc_rpm_smd_qol-v1-3-6dc5b5e97626@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysinterconnect: qcom: implement get_bw with rpmh_readNeil Armstrong
Since we can actually read back the APPS rpmh interconnect BCM votes we can actually implement the get_bw() callback and provide a coherent average and peak bandwidth at probe time. The benefits of that are: - keep disabled BCMs disabled - avoid voting unused BCMs to INT_MAX If the interconnects are correctly described for a platform, all the required BCMs would be voted to the maximum bandwidth until sync_state is reached. Since we only get the BCM vote, we need to redistribute the vote values to the associated nodes. The initial BCM votes are read back at probe time in order to be ready when the get_bw() is called when a node is added. Tested-by: Georgi Djakov <djakov@kernel.org> #db845c Signed-off-by: Neil Armstrong <neil.armstrong@linaro.org> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Link: https://patch.msgid.link/20260909-topic-sm8x50-icc-read-rpmh-v3-1-2bc65d7b25e5@linaro.org Signed-off-by: Georgi Djakov <djakov@kernel.org>
14 daysinterconnect: exynos: Remove the PM QoS request on unbindMyeonghun Pak
The Exynos interconnect provider adds a minimum-frequency PM QoS request to its parent devfreq device. The request is embedded in the provider's device-managed private data. The probe error path removes the request, but the remove callback does not. Unbinding the provider while the parent device remains registered therefore frees the private data with the request still on the parent's QoS list. The child's device cleanup does not remove requests registered against its parent. Remove the request after deregistering the provider, before the private data is released. This issue was identified during our ongoing static-analysis research while reviewing kernel code. Fixes: 2f95b9d5cf0b ("interconnect: Add generic interconnect driver for Exynos SoCs") Co-developed-by: Ijae Kim <ae878000@gmail.com> Signed-off-by: Ijae Kim <ae878000@gmail.com> Signed-off-by: Myeonghun Pak <mhun512@gmail.com> Link: https://patch.msgid.link/20260911031750.8182-1-mhun512@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-16Merge branch 'icc-kuno' into icc-nextGeorgi Djakov
Add interconnect (NoC) provider driver for the Qualcomm Kuno SoC. * icc-kuno dt-bindings: interconnect: Add Qualcomm Kuno NoC interconnect: qcom: Add Kuno interconnect provider driver Link: https://patch.msgid.link/20260827-kuno-soc-support-v5-0-6d47636a8f09@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-16Revert "interconnect: qcom: x1e80100: enable QoS configuration"Marc Zyngier
This reverts commit 5a8b2cc36e796d595c9d97eab23c2be22805cc1c. Since the introduction of this QoS change, X1 machines based on the Purwa SoC (such as the Lenovo Mini X) take a hard reset at boot time. Other machines based on Hamoa (such as the X1E001DE Snapdragon Devkit) end-up with a similar hard reset while under load, most likely due to the PCIe ports being starved of traffic. Revert the whole thing until someone figures out what magic parameters allow QoS to work in a sensible manner, as a usable machine is somehow preferable to one that crashes efficiently. Link: https://lore.kernel.org/all/apgeBNSiP01YfNcs@google.com Link: https://lore.kernel.org/all/86ld9d4h4n.wl-maz@kernel.org Signed-off-by: Marc Zyngier <maz@kernel.org> Cc: Georgi Djakov <djakov@kernel.org> Cc: Raviteja Laggyshetty <raviteja.laggyshetty@oss.qualcomm.com> Cc: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Cc: Abel Vesa <abelvesa@kernel.org> Cc: Bjorn Andersson <andersson@kernel.org> Cc: Mostafa Saleh <smostafa@google.com> Cc: Thorsten Leemhuis <regressions@leemhuis.info> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Reviewed-by: Mostafa Saleh <smostafa@google.com> Tested-by: Mostafa Saleh <smostafa@google.com> Closes: https://krzk.eu/#/builders/102/builds/70/steps/23/logs/warnings__3_ Link: https://patch.msgid.link/20260915123703.3228329-1-maz@kernel.org Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-16interconnect: Add kernel-doc for devm_of_icc_get(), icc_enable() and ↵Karl Mehltretter
icc_disable() Documentation/driver-api/interconnect.rst lists devm_of_icc_get(), icc_enable() and icc_disable() in its kernel-doc directive for drivers/interconnect/core.c, but none of the three functions has a kernel-doc comment. The directive silently produces nothing for them, so the consumer API section of the rendered documentation never shows them. Add kernel-doc comments in the style of the neighbouring functions. Assisted-by: LLM Signed-off-by: Karl Mehltretter <kmehltretter@gmail.com> Link: https://patch.msgid.link/20260912115456.89169-1-kmehltretter@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-16interconnect: qcom: icc-rpm: Don't skip small votesKonrad Dybcio
Small votes (such as 1 or 100) can end up being rounded down to 0 midway through the calculations. Use qcom_bw_div() when dividing to prevent that. Assisted-by: LLM Fixes: 30c8fa3ec61a ("interconnect: qcom: Add MSM8916 interconnect provider driver") Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Link: https://patch.msgid.link/20260909-topic-icc_rpm_smd_qol-v1-2-6dc5b5e97626@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-16interconnect: qcom: Move bcm_div() to common codeKonrad Dybcio
This helper is useful for both SMD_RPM and RPMH code, extract it. Assisted-by: LLM Signed-off-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Link: https://patch.msgid.link/20260909-topic-icc_rpm_smd_qol-v1-1-6dc5b5e97626@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-09interconnect: fix typo "commited" in commentsHemanth Selam
Correct "commited" to "committed", reported by scripts/checkpatch.pl using the misspelling list in scripts/spelling.txt. Only touches comments, no code changes. Assisted-by: Cursor:claude-opus-5 Signed-off-by: Hemanth Selam <hemanth.selam@gmail.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260904111632.29262-1-hemanth.selam@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-09interconnect: qcom: Constify struct qcom_icc_qosboxChristophe JAILLET
'struct qcom_icc_qosbox' are not modified in these drivers. Constifying these structures moves some data to a read-only section, so increases overall security. On a x86_64, with allmodconfig: Before: ====== text data bss dec hex filename 7807 18632 0 26439 6747 drivers/interconnect/qcom/eliza.o 10027 20176 0 30203 75fb drivers/interconnect/qcom/milos.o 8768 22640 0 31408 7ab0 drivers/interconnect/qcom/sm6350.o 7626 22000 0 29626 73ba drivers/interconnect/qcom/sm8650.o After: ===== text data bss dec hex filename 8095 18344 0 26439 6747 drivers/interconnect/qcom/eliza.o 11079 19120 0 30199 75f7 drivers/interconnect/qcom/milos.o 9668 21744 0 31412 7ab4 drivers/interconnect/qcom/sm6350.o 8746 20880 0 29626 73ba drivers/interconnect/qcom/sm8650.o Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Link: https://patch.msgid.link/f9d9ab49e961b3bf3f9bad39f22cf8b06afb2381.1786554463.git.christophe.jaillet@wanadoo.fr Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-09interconnect: qcom: Add Kuno interconnect provider driverHardeep Sharma
Add the interconnect provider driver for the Qualcomm Kuno SoC to enable bus bandwidth voting on the platform. This includes the aggre-noc and cnoc-main NoC instances with their associated BCM nodes. Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Signed-off-by: Hardeep Sharma <hardeep.sharma@oss.qualcomm.com> Reviewed-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Link: https://patch.msgid.link/20260827-kuno-soc-support-v5-7-6d47636a8f09@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-08interconnect: qcom: Add missing depends to Hawi, Maili, Nord and ShikraKrzysztof Kozlowski
Hawi, Maili, Nord and Shikra were added in parallel or after commit 5b696f065843 ("interconnect: qcom: Restrict drivers per ARM/ARM64"), thus they missed the update restricting them per architecture (with the same rationale). Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260720-interconnect-qcom-clean-arm64-v3-2-942e8ee7ebda@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-08interconnect: qcom: Make important drivers defaultKrzysztof Kozlowski
The interconnect drivers for Qualcomm SoC Network-on-Chip are covering a basic or fundamental SoC feature: bandwidth management between internal SoC blocks. SoC can boot without these, but power management or performance will be affected. These drivers do not represent any sort of buses visible to the board designers/configurators, thus they should be always enabled, regardless how SoC is used in the final board. Kernel configuration should not ask users choice of drivers when that choice is obvious and known to the developers that answer should be 'yes' or 'module'. Switch all almost Qualcomm interconnect drivers to a default 'yes' for ARCH_QCOM. This has impact: 1. arm64 defconfig: a. Enable as built-in INTERCONNECT_QCOM_HAWI, INTERCONNECT_QCOM_NORD, INTERCONNECT_QCOM_SDM660, INTERCONNECT_QCOM_SDM670, INTERCONNECT_QCOM_SM7150 and INTERCONNECT_QCOM_SAR2130P, which were not selected before but should be, because these platforms need them anyway for proper functioning. b. Switch to built-in from a module INTERCONNECT_QCOM_QCS404 and INTERCONNECT_QCOM_MSM8916, which as modules would not make the platform bootable in most cases, and INTERCONNECT_QCOM_OSM_L3, which when module might slow down boot considerably by having caches running at slow speed. 2. arm qcom_defconfig: Switch to built-in from a module INTERCONNECT_QCOM_RPMH, INTERCONNECT_QCOM_SMD_RPM, INTERCONNECT_QCOM_BCM_VOTER, INTERCONNECT_QCOM_MSM8974, INTERCONNECT_QCOM_SDX55, which as modules would not make the platform bootable in most cases. 3. arm multi_v7 defconfig: Enable drivers necessary to boot ARM 32-bit platforms, which are already enabled on qcom_defconfig: a. Enable as built-in INTERCONNECT_QCOM_MSM8974. b. Enable as modules (other dependencies prevent from built-in) INTERCONNECT_QCOM_RPMH, INTERCONNECT_QCOM_BCM_VOTER and INTERCONNECT_QCOM_SDX55. 4. COMPILE_TEST builds: Enable by default all drivers for arm or arm64 builds, whenever ARCH_QCOM is selected. This has impact on build time and feels logical, because if one selects ARCH_QCOM then probably by default wants to build test it entirely. Kernels with COMPILE_TEST are not supposed to be used for booting. Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260720-interconnect-qcom-clean-arm64-v3-1-942e8ee7ebda@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-09-08interconnect: qcom: Enable Shikra interconnect driver by default for ARCH_QCOMRaviteja Laggyshetty
Interconnect drivers provide fundamental NoC bandwidth management required for correct system behavior. Although systems can boot without them, power and performance are impacted. These drivers need to enabled irresepective of the board variant, design or configuration. Enable the Shikra interconnect driver by default on ARCH_QCOM by setting "default ARCH_QCOM". Signed-off-by: Raviteja Laggyshetty <raviteja.laggyshetty@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260526-shikra_icc_kconfig-v1-1-c589db2d023c@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-08-19Merge tag 'icc-7.3-rc1' of ↵Greg Kroah-Hartman
ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/djakov/icc into char-misc-next Georgi writes: This pull request contains the following interconnect updates for the 7.3-rc1 merge window: - New driver for Maili SoC - Add support for QoS on the SC8280XP SoC - Add support for QoS on the x1e80100 SoC - Add EPSS L3 scaling support for Shikra SoC - Add COMPILE_TEST support for some platforms - Misc tiny improvements and fixes Signed-off-by: Georgi Djakov <djakov@kernel.org> * tag 'icc-7.3-rc1' of ssh://gitolite.kernel.org/pub/scm/linux/kernel/git/djakov/icc: interconnect: Fix use after free in icc_get() and of_icc_get_by_index() interconnect: debugfs-client: add NULL check for platform_device_alloc interconnect: qcom: simplify allocation interconnect: qcom: add COMPILE_TEST interconnect: qcom: add Maili interconnect provider driver dt-bindings: interconnect: qcom: document the RPMh Network-On-Chip interconnect in Maili SoC interconnect: qcom: Add EPSS L3 scaling support for Shikra SoC dt-bindings: interconnect: qcom,osm-l3: Add EPSS L3 DT binding for Qualcomm Shikra SoC interconnect: qcom: x1e80100: enable QoS configuration dt-bindings: interconnect: qcom,x1e80100-rpmh: add clocks property to enable QoS interconnect: qcom: sc8280xp: Enable QoS configuration dt-bindings: interconnect: qcom,sc8280xp-rpmh: Add reg and clocks for QoS
2026-07-22Merge branch 'icc-misc' into icc-nextGeorgi Djakov
* icc-misc interconnect: qcom: add COMPILE_TEST interconnect: qcom: simplify allocation interconnect: debugfs-client: add NULL check for platform_device_alloc interconnect: Fix use after free in icc_get() and of_icc_get_by_index() Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-22Merge branch 'icc-maili' into icc-nextGeorgi Djakov
Add interconnect bindings and RPMh-based interconnect driver support for the upcoming Qualcomm Maili SoC. * icc-maili dt-bindings: interconnect: qcom: document the RPMh Network-On-Chip interconnect in Maili SoC interconnect: qcom: add Maili interconnect provider driver Link: https://patch.msgid.link/20260622-maili_icc-v2-0-18b5ac08c04f@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-22Merge branch 'icc-shikra' into icc-nextGeorgi Djakov
Add Epoch Subsystem (EPSS) L3 scaling support on Qualcomm Shikra SoC. EPSS hardware on Shikra is similar to other SoCs but supports only twelve L3 frequency entries (LUT). Reading the LUT beyond supported frequencies can expose incorrect frequencies. Introduce new compatible to represent this constrained variant of EPSS. * icc-shikra dt-bindings: interconnect: qcom,osm-l3: Add EPSS L3 DT binding for Qualcomm Shikra SoC interconnect: qcom: Add EPSS L3 scaling support for Shikra SoC Link: https://patch.msgid.link/20260603-shikra_epss_l3-v3-0-3c2e0b796e78@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-22interconnect: Fix use after free in icc_get() and of_icc_get_by_index()Kuan-Wei Chiu
In of_icc_get_by_index() and icc_get(), if the dynamic allocation for path->name fails via kasprintf(), the error handling path directly calls kfree(path) to free the path object and returns an error. However, prior to this point, path_find() calls path_init(), which already links the path's requests into the req_list of the respective interconnect nodes via hlist_add_head(). Directly invoking kfree(path) leaves dangling pointers in the hlist. A subsequent call to icc_get() or icc_set_bw() will traverse or modify these corrupted lists, triggering a slab use afterfree. KASAN report showing the vulnerability when reproducing via debugfs: BUG: KASAN: slab-use-after-free in path_find+0x6f8/0xcfc Write of size 8 at addr fff000000d43f748 by task sh/1 ... Call trace: kasan_report+0xac/0xfc path_find+0x6f8/0xcfc icc_get+0x148/0x380 icc_get_set+0xf8/0x2d0 ... Freed by task 1: kfree+0x1a0/0x4a4 icc_get+0x2cc/0x380 icc_get_set+0xf8/0x2d0 Fix this by replacing kfree(path) with the proper teardown function, icc_put(path), which safely removes the requests from the req_list using hlist_del() and drops the provider usage references before freeing the memory. Additionally, in icc_get(), ensure that the icc_lock mutex is released prior to calling icc_put(path) to avoid a deadlock, as icc_put() internally acquires the same lock. Fixes: 3791163602f7 ("interconnect: Handle memory allocation errors") Cc: stable@vger.kernel.org Signed-off-by: Kuan-Wei Chiu <visitorckw@gmail.com> Link: https://patch.msgid.link/20260416190840.1753468-1-visitorckw@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-22interconnect: debugfs-client: add NULL check for platform_device_allocLi Jun
platform_device_alloc() may return NULL when memory allocation fails. Add proper error handling to prevent a potential NULL pointer dereference when calling platform_device_add() with a NULL pointer. Signed-off-by: Li Jun <lijun01@kylinos.cn> Link: https://patch.msgid.link/20260602103120.2966458-1-lijun01@kylinos.cn Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-22interconnect: qcom: simplify allocationRosen Penev
Use a flexible array member to reduce allocation by 1. Add __counted_by for extra runtime analysis. Move counting variable assignment after allocation before any array access. Signed-off-by: Rosen Penev <rosenp@gmail.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260609222454.37352-1-rosenp@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-22interconnect: qcom: add COMPILE_TESTRosen Penev
Allow non-ARM to build this. Signed-off-by: Rosen Penev <rosenp@gmail.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Link: https://patch.msgid.link/20260608050542.6167-1-rosenp@gmail.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-07interconnect: qcom: add Maili interconnect provider driverRaviteja Laggyshetty
Add driver for the Qualcomm interconnect buses found in Maili based platforms. The topology consists of several NoCs that are controlled by a remote processor that collects the aggregated bandwidth for each master-slave pairs. Co-developed-by: Odelu Kukatla <odelu.kukatla@oss.qualcomm.com> Signed-off-by: Odelu Kukatla <odelu.kukatla@oss.qualcomm.com> Signed-off-by: Raviteja Laggyshetty <raviteja.laggyshetty@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260622-maili_icc-v2-2-18b5ac08c04f@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-07interconnect: qcom: Add EPSS L3 scaling support for Shikra SoCRaviteja Laggyshetty
Add Epoch Subsystem (EPSS) L3 interconnect provider support on Qualcomm Shikra SoC. The EPSS L3 block on Shikra SoC is similar to existing Qualcomm EPSS/OSM L3 providers, but supports only up to 12 frequency lookup table entries. Reading beyond the supported LUT entries can expose incorrect frequencies. Add shikra-specific EPSS descriptor shikra_epss_l3_perf_state that reuses existing EPSS configuration with appropriate LUT entries limit. Co-developed-by: Odelu Kukatla <odelu.kukatla@oss.qualcomm.com> Signed-off-by: Odelu Kukatla <odelu.kukatla@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Signed-off-by: Raviteja Laggyshetty <raviteja.laggyshetty@oss.qualcomm.com> Link: https://patch.msgid.link/20260603-shikra_epss_l3-v3-2-3c2e0b796e78@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-07Merge branch 'icc-x1e80100' into icc-nextGeorgi Djakov
This series enables QoS configuration for QNOC type device which can be found on X1E80100 platform. It enables QoS configuration for master ports with predefined priority and urgency forwarding. This helps in prioritizing the traffic originating from different interconnect masters at NOC (Network On Chip). The system may function normally without this feature. However, enabling QoS helps optimize latency and bandwidth across subsystems like CPU, GPU, and multimedia engines, which becomes important in high-throughput scenarios. This is a feature aimed at performance enhancement to improve system performance under concurrent workloads. * icc-x1e80100 dt-bindings: interconnect: qcom,x1e80100-rpmh: add clocks property to enable QoS interconnect: qcom: x1e80100: enable QoS configuration Link: https://patch.msgid.link/20260527-x1e80100_qos-v2-0-305c6539e6d2@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-07interconnect: qcom: x1e80100: enable QoS configurationRaviteja Laggyshetty
Enable QoS configuration for master ports with predefined priority and urgency forwarding. Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Signed-off-by: Raviteja Laggyshetty <raviteja.laggyshetty@oss.qualcomm.com> Link: https://patch.msgid.link/20260527-x1e80100_qos-v2-2-305c6539e6d2@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-07interconnect: qcom: sc8280xp: Enable QoS configurationXilin Wu
Enable static QoS configuration for SC8280XP NoC master ports with predefined priority and urgency forwarding values. Add the QoS box data for the SC8280XP providers, add regmap configurations for the real NoCs, and mark only aggre1_noc and aggre2_noc as requiring clocks for QoS register access. Signed-off-by: Xilin Wu <sophon@radxa.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Link: https://patch.msgid.link/20260507-sc8280xp-qos-v1-2-15135858cd98@radxa.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-07-03Replace <linux/mod_devicetable.h> by more specific <linux/device-id/*.h> (c ↵Uwe Kleine-König (The Capable Hub)
files) Replace the #include of <linux/mod_devicetable.h> by the more specific <linux/device-id/*.h> where applicable. For most cases the include can be dropped completely, only a few drivers need one or two headers added. Acked-by: Danilo Krummrich <dakr@kernel.org> Acked-by: Takashi Sakamoto <o-takashi@sakamocchi.jp> Acked-by: Bjorn Helgaas <bhelgaas@google.com> Link: https://patch.msgid.link/1a3f2007c5c5dcf555c09a4035ce3ae8ef1b6c49.1782808461.git.u.kleine-koenig@baylibre.com Signed-off-by: Uwe Kleine-König (The Capable Hub) <u.kleine-koenig@baylibre.com>
2026-06-25Merge tag 'clk-for-linus' of ↵Linus Torvalds
git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux Pull clk updates from Stephen Boyd: "This is all clk driver updates. Mostly new SoC support for various Qualcomm chips and Canaan K230. Otherwise there's non-critical fixes and updates to clk data such as adding missing clks to existing drivers or marking clks critical. Nothing looks especially exciting" * tag 'clk-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/clk/linux: (106 commits) clk: qcom: regmap-phy-mux: Rework the implementation clk: qcom: a53: Corrected frequency multiplier for 1152MHz clk: qcom: camcc-milos: Declare icc path dependency for CAMSS_TOP_GDSC clk: qcom: gdsc: Support enabling interconnect path for power domain dt-bindings: clock: qcom,milos-camcc: Document interconnect path interconnect: Add devm_of_icc_get_by_index() as exported API for users clk: qcom: camcc-x1p42100: Add support for camera clock controller clk: qcom: camcc-x1e80100: Add support for camera QDSS debug clocks clk: qcom: videocc-x1p42100: Add support for video clock controller dt-bindings: clock: qcom: Add X1P42100 camera clock controller dt-bindings: clock: qcom: Add X1P42100 video clock controller clk: keystone: sci-clk: fix application of sizeof to pointer clk: keystone: don't cache clock rate clk: spacemit: k3: Add PCIe DBI clock dt-bindings: soc: spacemit: k3: Add PCIe DBI clock IDs clk: spacemit: k3: Fix PCIe clock register offset clk: spacemit: k3: Switch to pll2_d6 as parent for PCIe clock clk: at91: keep securam node alive while mapping it clk: samsung: exynos990: Fix PERIC0/1 USI clock types clk: renesas: r9a08g045: Drop unused pm_domain header file ...
2026-06-06interconnect: Add devm_of_icc_get_by_index() as exported API for usersLuca Weiss
Users can use devm version of of_icc_get_by_index() to benefit from automatic resource release. Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Signed-off-by: Luca Weiss <luca.weiss@fairphone.com> Acked-by: Georgi Djakov <djakov@kernel.org> Link: https://lore.kernel.org/r/20260501-milos-camcc-icc-v2-1-bb83c1256cc3@fairphone.com Signed-off-by: Bjorn Andersson <andersson@kernel.org>
2026-05-22Merge branch 'icc-misc' into icc-nextGeorgi Djakov
* icc-misc interconnect: Do not create empty devres on missing interconnects interconnect: Move MODULE_DEVICE_TABLE next to the table itself dt-bindings: interconnect: qcom,sdm660: Disallow clocks when appropriate dt-bindings: interconnect: qcom,sm6115: Drop incorrect children if:then: block dt-bindings: interconnect: qcom,sm6115: Restrict children and clocks interconnect: qcom: Fix indentation interconnect: qcom: Restrict drivers per ARM/ARM64 Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-21Merge branch 'icc-hawi' into icc-nextGeorgi Djakov
Add interconnect bindings and RPMh-based interconnect driver support for the upcoming Qualcomm Hawi SoC. * icc-hawi dt-bindings: interconnect: qcom-bwmon: Add Hawi cpu-bwmon compatible dt-bindings: interconnect: qcom-bwmon: Add Hawi llcc-bwmon compatible dt-bindings: interconnect: qcom: document the RPMh NoC for Hawi SoC interconnect: qcom: add Hawi interconnect provider driver Link: https://patch.msgid.link/20260506-icc-hawi-v4-0-35447fdc482b@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-21interconnect: qcom: add Hawi interconnect provider driverVivek Aknurwar
Add driver for the Qualcomm interconnect buses found in Hawi based platforms. The topology consists of several NoCs that are controlled by a remote processor that collects the aggregated bandwidth for each master-slave pair. Reviewed-by: Mike Tipton <mike.tipton@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com> Signed-off-by: Vivek Aknurwar <vivek.aknurwar@oss.qualcomm.com> Link: https://patch.msgid.link/20260506-icc-hawi-v4-2-35447fdc482b@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-21Merge branch 'icc-eliza' into icc-nextGeorgi Djakov
Add the missing Eliza SDCC1 interconnect slave ID and provider node. The Eliza interconnect binding and provider already describe SDCC2, but the matching SDCC1 CNOC CFG slave was left out. Add the binding constant and the provider node so consumers can describe SDCC1 bandwidth paths. The provider change also adds qhs_sdc1 to qsm_cfg and bcm_cn0, and updates the qsm_cfg link count and bcm_cn0 node count. * icc-eliza dt-bindings: interconnect: qcom,eliza-rpmh: Add SDCC1 slave interconnect: qcom: eliza: Add SDCC1 slave node Link: https://patch.msgid.link/20260514-eliza-interconnect-add-missing-sdcc1-slave-node-v2-0-13c03bc890cb@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-21Merge branch 'icc-nord' into icc-nextGeorgi Djakov
This adds RPMh-based interconnect support for the Qualcomm Nord SoC. The Nord SoC features a rich Network-on-Chip topology comprising 19 NoCs including aggregate NoCs, a high-speed configuration NoC (HSCNOC), a multimedia NoC, four NSP data NoCs for AI/ML workloads, PCIe inbound and outbound NoCs, a system NoC, and virtual clock/MC nodes. Bandwidth requests are communicated to the RPMh hardware through Bus Clock Manager (BCM) resources via the Resource State Coordinator (RSC). * icc-nord dt-bindings: interconnect: Document RPMh Network-On-Chip for Qualcomm Nord SoC interconnect: qcom: Add interconnect provider driver for Nord SoC Link: https://patch.msgid.link/20260510020607.1129773-1-shengchao.guo@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-15interconnect: qcom: eliza: Add SDCC1 slave nodeAbel Vesa
The Eliza interconnect provider is missing the SDCC1 CNOC CFG slave node. Add qhs_sdc1 to the provider node table so SDCC1 interconnect paths can resolve to a provider node. Hook qhs_sdc1 up to qsm_cfg and CN0, and bump the corresponding qsm_cfg.num_links and bcm_cn0.num_nodes counts. Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260514-eliza-interconnect-add-missing-sdcc1-slave-node-v2-2-13c03bc890cb@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-10interconnect: qcom: Restrict drivers per ARM/ARM64Krzysztof Kozlowski
There is no point to allow selecting core SoC drivers like interconnects for Qualcomm ARMv7 SoCs when building ARM64 kernel, and vice versa. This makes kernel configuration more difficult as many do not remember the Qualcomm SoCs model names/numbers and their properties like architecture. No features should be lost because: 1. There won't be a single image for ARMv7 and ARMv8/9 SoCs. 2. Newer ARMv8/9 SoCs won't be running in arm32 emulation mode. Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260428-interconnect-qcom-clean-arm64-v1-2-e6bc3f7832db@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-10interconnect: qcom: Fix indentationKrzysztof Kozlowski
KConfig entries should be indented starting with one tab, so replace spaces with it. Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com> Link: https://patch.msgid.link/20260428-interconnect-qcom-clean-arm64-v1-1-e6bc3f7832db@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>
2026-05-10interconnect: Move MODULE_DEVICE_TABLE next to the table itselfKrzysztof Kozlowski
By convention MODULE_DEVICE_TABLE() immediately follows the ID table it exports, because this is easier to read and verify. It also makes more sense since #ifdef for ACPI or OF could hide both of them. Most of the privers already have this correctly placed, so adjust the missing ones. No functional impact. Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@oss.qualcomm.com> Reviewed-by: Dmitry Baryshkov <dmitry.baryshkov@oss.qualcomm.com> Reviewed-by: Konrad Dybcio <konrad.dybcio@oss.qualcomm.com> Link: https://patch.msgid.link/20260505102854.186925-2-krzysztof.kozlowski@oss.qualcomm.com Signed-off-by: Georgi Djakov <djakov@kernel.org>