aboutsummaryrefslogtreecommitdiff
path: root/drivers
AgeCommit message (Collapse)AuthorFilesLines
2024-08-15ata: libahci_platform: Simplify code with for_each_child_of_node_scoped()Zhang Zekun1-14/+5
for_each_child_of_node_scoped() can put the device_node automatically. So let's use it to make the code simpler by avoiding the need to explicitly call of_node_put(). Signed-off-by: Zhang Zekun <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Niklas Cassel <[email protected]>
2024-08-15net: hns3: use correct release function during uninitializationPeiyang Wang1-1/+1
pci_request_regions is called to apply for PCI I/O and memory resources when the driver is initialized, Therefore, when the driver is uninstalled, pci_release_regions should be used to release PCI I/O and memory resources instead of pci_release_mem_regions is used to release memory reasouces only. Signed-off-by: Peiyang Wang <[email protected]> Signed-off-by: Jijie Shao <[email protected]> Signed-off-by: Paolo Abeni <[email protected]>
2024-08-15net: hns3: void array out of bound when loop tnl_numPeiyang Wang1-3/+3
When query reg inf of SSU, it loops tnl_num times. However, tnl_num comes from hardware and the length of array is a fixed value. To void array out of bound, make sure the loop time is not greater than the length of array Signed-off-by: Peiyang Wang <[email protected]> Signed-off-by: Jijie Shao <[email protected]> Signed-off-by: Paolo Abeni <[email protected]>
2024-08-15net: hns3: fix a deadlock problem when config TC during resettingJie Wang1-0/+3
When config TC during the reset process, may cause a deadlock, the flow is as below: pf reset start │ ▼ ...... setup tc │ │ ▼ ▼ DOWN: napi_disable() napi_disable()(skip) │ │ │ ▼ ▼ ...... ...... │ │ ▼ │ napi_enable() │ ▼ UINIT: netif_napi_del() │ ▼ ...... │ ▼ INIT: netif_napi_add() │ ▼ ...... global reset start │ │ ▼ ▼ UP: napi_enable()(skip) ...... │ │ ▼ ▼ ...... napi_disable() In reset process, the driver will DOWN the port and then UINIT, in this case, the setup tc process will UP the port before UINIT, so cause the problem. Adds a DOWN process in UINIT to fix it. Fixes: bb6b94a896d4 ("net: hns3: Add reset interface implementation in client") Signed-off-by: Jie Wang <[email protected]> Signed-off-by: Jijie Shao <[email protected]> Signed-off-by: Paolo Abeni <[email protected]>
2024-08-15net: hns3: use the user's cfg after resetPeiyang Wang2-6/+21
Consider the followed case that the user change speed and reset the net interface. Before the hw change speed successfully, the driver get old old speed from hw by timer task. After reset, the previous speed is config to hw. As a result, the new speed is configed successfully but lost after PF reset. The followed pictured shows more dirrectly. +------+ +----+ +----+ | USER | | PF | | HW | +---+--+ +-+--+ +-+--+ | ethtool -s 100G | | +------------------>| set speed 100G | | +--------------------->| | | set successfully | | |<---------------------+---+ | |query cfg (timer task)| | | +--------------------->| | handle speed | | return 200G | | changing event | ethtool --reset |<---------------------+ | (100G) +------------------>| cfg previous speed |<--+ | | after reset (200G) | | +--------------------->| | | +---+ | |query cfg (timer task)| | | +--------------------->| | handle speed | | return 100G | | changing event | |<---------------------+ | (200G) | | |<--+ | |query cfg (timer task)| | +--------------------->| | | return 200G | | |<---------------------+ | | | v v v This patch save new speed if hw change speed successfully, which will be used after reset successfully. Fixes: 2d03eacc0b7e ("net: hns3: Only update mac configuation when necessary") Signed-off-by: Peiyang Wang <[email protected]> Signed-off-by: Jijie Shao <[email protected]> Signed-off-by: Paolo Abeni <[email protected]>
2024-08-15net: hns3: fix wrong use of semaphore upJie Wang2-4/+4
Currently, if hns3 PF or VF FLR reset failed after five times retry, the reset done process will directly release the semaphore which has already released in hclge_reset_prepare_general. This will cause down operation fail. So this patch fixes it by adding reset state judgement. The up operation is only called after successful PF FLR reset. Fixes: 8627bdedc435 ("net: hns3: refactor the precedure of PF FLR") Fixes: f28368bb4542 ("net: hns3: refactor the procedure of VF FLR") Signed-off-by: Jie Wang <[email protected]> Signed-off-by: Jijie Shao <[email protected]> Signed-off-by: Paolo Abeni <[email protected]>
2024-08-15pse-core: Conditionally set current limit during PI regulator registrationOleksij Rempel1-3/+8
Fix an issue where `devm_regulator_register()` would fail for PSE controllers that do not support current limit control, such as simple GPIO-based controllers like the podl-pse-regulator. The `REGULATOR_CHANGE_CURRENT` flag and `max_uA` constraint are now conditionally set only if the `pi_set_current_limit` operation is supported. This change prevents the regulator registration routine from attempting to call `pse_pi_set_current_limit()`, which would return `-EOPNOTSUPP` and cause the registration to fail. Fixes: 4a83abcef5f4f ("net: pse-pd: Add new power limit get and set c33 features") Signed-off-by: Oleksij Rempel <[email protected]> Reviewed-by: Kory Maincent <[email protected]> Tested-by: Kyle Swenson <[email protected]> Link: https://patch.msgid.link/[email protected] Signed-off-by: Paolo Abeni <[email protected]>
2024-08-15pmdomain: Merge branch fixes into nextUlf Hansson1-2/+3
Merge the pmdomain fixes for v6.11-rc[n] into the next branch, to allow them to get tested together with the new changes that are targeted for v6.12. Signed-off-by: Ulf Hansson <[email protected]>
2024-08-15pmdomain: imx: wait SSAR when i.MX93 power domain onPeng Fan1-2/+3
With "quiet" set in bootargs, there is power domain failure: "imx93_power_domain 44462400.power-domain: pd_off timeout: name: 44462400.power-domain, stat: 4" The current power on opertation takes ISO state as power on finished flag, but it is wrong. Before powering on operation really finishes, powering off comes and powering off will never finish because the last powering on still not finishes, so the following powering off actually not trigger hardware state machine to run. SSAR is the last step when powering on a domain, so need to wait SSAR done when powering on. Since EdgeLock Enclave(ELE) handshake is involved in the flow, enlarge the waiting time to 10ms for both on and off to avoid timeout. Cc: [email protected] Fixes: 0a0f7cc25d4a ("soc: imx: add i.MX93 SRC power domain driver") Reviewed-by: Jacky Bai <[email protected]> Signed-off-by: Peng Fan <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Ulf Hansson <[email protected]>
2024-08-15drm/rockchip: inno-hdmi: Fix infoframe uploadAlex Bee1-3/+1
HDMI analyser shows that the AVI infoframe is no being longer send. The switch to the HDMI connector api should have used the frame content which is now given in the buffer parameter, but instead still uses the (now) empty and superfluous packed_frame variable. Fix it. Fixes: 65548c8ff0ab ("drm/rockchip: inno_hdmi: Switch to HDMI connector") Signed-off-by: Alex Bee <[email protected]> Acked-by: Maxime Ripard <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2024-08-15net: thunder_bgx: Fix netdev structure allocationMarc Zyngier1-9/+21
Commit 94833addfaba ("net: thunderx: Unembed netdev structure") had a go at dynamically allocating the netdev structures for the thunderx_bgx driver. This change results in my ThunderX box catching fire (to be fair, it is what it does best). The issues with this change are that: - bgx_lmac_enable() is called *after* bgx_acpi_register_phy() and bgx_init_of_phy(), both expecting netdev to be a valid pointer. - bgx_init_of_phy() populates the MAC addresses for *all* LMACs attached to a given BGX instance, and thus needs netdev for each of them to have been allocated. There is a few things to be said about how the driver mixes LMAC and BGX states which leads to this sorry state, but that's beside the point. To address this, go back to a situation where all netdev structures are allocated before the driver starts relying on them, and move the freeing of these structures to driver removal. Someone brave enough can always go and restructure the driver if they want. Fixes: 94833addfaba ("net: thunderx: Unembed netdev structure") Signed-off-by: Marc Zyngier <[email protected]> Cc: Breno Leitao <[email protected]> Cc: Sunil Goutham <[email protected]> Cc: "David S. Miller" <[email protected]> Cc: Eric Dumazet <[email protected]> Cc: Jakub Kicinski <[email protected]> Cc: Paolo Abeni <[email protected]> Reviewed-by: Simon Horman <[email protected]> Reviewed-by: Breno Leitao <[email protected]> Link: https://patch.msgid.link/[email protected] Signed-off-by: Paolo Abeni <[email protected]>
2024-08-15drm/rockchip: dw_hdmi: Drop superfluous assignments of mpll_cfg, cur_ctr and ↵Cristian Ciocaltea1-6/+0
phy_config The mpll_cfg, cur_ctr and phy_config members in struct dw_hdmi_plat_data are only used to configure the Synopsys PHYs supported internally by DW HDMI transmitter driver (gpu/drm/bridge/synopsys/dw-hdmi.c), via hdmi_phy_configure_dwc_hdmi_3d_tx(), which is further invoked from dw_hdmi_phy_init(). This is part of the internal dw_hdmi_synopsys_phy_ops struct, managed within dw_hdmi_detect_phy(). To handle vendor PHYs, DW HDMI driver doesn't make use of the internal PHY ops and, instead, relies on the glue layer to provide the phy_ops and phy_name members of struct dw_hdmi_plat_data. Drop the unnecessary assignments of DW internal PHY related members from structs rk3228_hdmi_drv_data and rk3328_hdmi_drv_data, since both set the phy_force_vendor flag and correctly provide the expected vendor PHY data. Signed-off-by: Cristian Ciocaltea <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/20240813-dw-hdmi-rockchip-cleanup-v1-4-b3e73b5f4fd6@collabora.com
2024-08-15drm/rockchip: dw_hdmi: Use devm_regulator_get_enable()Cristian Ciocaltea1-28/+6
The regulators are only enabled at bind() and disabled at unbind(), hence replace the boilerplate code by making use of devm_regulator_get_enable() helper. Signed-off-by: Cristian Ciocaltea <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/20240813-dw-hdmi-rockchip-cleanup-v1-3-b3e73b5f4fd6@collabora.com
2024-08-15drm/rockchip: dw_hdmi: Simplify clock handlingCristian Ciocaltea1-24/+14
Make use of devm_clk_get_optional_enabled() to replace devm_clk_get() and clk_prepare_enable() for ref_clk and drop the now unnecessary calls to clk_disable_unprepare(). Additionally, use devm_clk_get_optional() helper for grf_clk to replace the open coding call to devm_clk_get() followed by the -ENOENT test. Signed-off-by: Cristian Ciocaltea <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/20240813-dw-hdmi-rockchip-cleanup-v1-2-b3e73b5f4fd6@collabora.com
2024-08-15drm/rockchip: dw_hdmi: Use modern drm_device based loggingCristian Ciocaltea1-13/+11
Prefer drm_{err|info|dbg}() over deprecated DRM_DEV_{ERROR|INFO|DEBUG}() logging macros. Conversion done with the help of the following semantic patch, followed by a few minor indentation adjustments: @@ identifier T; @@ ( -DRM_DEV_ERROR(T->dev, +drm_err(T, ...) | -DRM_DEV_INFO(T->dev, +drm_info(T, ...) | -DRM_DEV_DEBUG(T->dev, +drm_dbg(T, ...) ) Signed-off-by: Cristian Ciocaltea <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/20240813-dw-hdmi-rockchip-cleanup-v1-1-b3e73b5f4fd6@collabora.com
2024-08-15drm/rockchip: vop: enable VOP_FEATURE_INTERNAL_RGB on RK3066Val Packett1-0/+1
The RK3066 does have RGB display output, so it should be marked as such. Fixes: f4a6de855eae ("drm: rockchip: vop: add rk3066 vop definitions") Cc: [email protected] Signed-off-by: Val Packett <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2024-08-15drm/rockchip: vop: clear DMA stop bit on RK3066Val Packett3-0/+6
The RK3066 VOP sets a dma_stop bit when it's done scanning out a frame and needs the driver to acknowledge that by clearing the bit. Unless we clear it "between" frames, the RGB output only shows noise instead of the picture. atomic_flush is the place for it that least affects other code (doing it on vblank would require converting all other usages of the reg_lock to spin_(un)lock_irq, which would affect performance for everyone). This seems to be a redundant synchronization mechanism that was removed in later iterations of the VOP hardware block. Fixes: f4a6de855eae ("drm: rockchip: vop: add rk3066 vop definitions") Cc: [email protected] Signed-off-by: Val Packett <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
2024-08-15drm/rockchip: Explicitly include bits headerCristian Ciocaltea1-1/+2
Driver makes use of the BIT() macro, but relies on the bits header being implicitly included. Explicitly pull the header in to avoid potential build failures in some configurations. While at it, reorder include directives alphabetically. Signed-off-by: Cristian Ciocaltea <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/20240807-b4-rk3588-bridge-upstream-v3-4-60d6bab0dc7c@collabora.com
2024-08-15drm/rockchip: Constify struct drm_encoder_helper_funcsChristophe JAILLET2-2/+2
'struct drm_encoder_helper_funcs' is not modified in these drivers. Constifying this structure moves some data to a read-only section, so increase overall security. On a x86_64, with allmodconfig: Before: ====== text data bss dec hex filename 7458 552 0 8010 1f4a drivers/gpu/drm/rockchip/analogix_dp-rockchip.o After: ===== text data bss dec hex filename 7578 424 0 8002 1f42 drivers/gpu/drm/rockchip/analogix_dp-rockchip.o Signed-off-by: Christophe JAILLET <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/128f9941aab3b1367eb7abca4ac26e2e5dd6ad21.1720903899.git.christophe.jaillet@wanadoo.fr
2024-08-15drm/rockchip: cdn-dp: Clean up a few logged messagesDragan Simic1-9/+7
Clean up a few logged messages, which were previously worded as rather incomplete sentences separated by periods. This was both a bit unreadable and grammatically incorrect, so convert them into partial sentences separated (or connected) by semicolons, together with some wording improvements. Signed-off-by: Dragan Simic <[email protected]> Reviewed-by: Andy Yan <[email protected]> Signed-off-by: Heiko Stuebner <[email protected]> Link: https://patchwork.freedesktop.org/patch/msgid/92db74a313547c087cc71059428698c4ec37a9ae.1720048818.git.dsimic@manjaro.org
2024-08-15md: convert comma to semicolonChen Ni1-1/+1
Replace a comma between expression statements by a semicolon. Fixes: 5e5702898e93 ("md/raid10: Handle read errors during recovery better.") Signed-off-by: Chen Ni <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Song Liu <[email protected]>
2024-08-15Revert "serial: 8250_omap: Set the console genpd always on if no console ↵Griffin Kroah-Hartman1-28/+5
suspend" This reverts commit 68e6939ea9ec3d6579eadeab16060339cdeaf940. Kevin reported that this causes a crash during suspend on platforms that dont use PM domains. Link: https://lore.kernel.org/r/[email protected] Cc: Thomas Richard <[email protected]> Fixes: 68e6939ea9ec ("serial: 8250_omap: Set the console genpd always on if no console suspend") Cc: stable <[email protected]> Reported-by: Kevin Hilman <[email protected]> Signed-off-by: Griffin Kroah-Hartman <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Greg Kroah-Hartman <[email protected]>
2024-08-14Merge tag 'wireless-2024-08-14' of ↵Jakub Kicinski10-17/+123
git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless Kalle Valo says: ==================== wireless fixes for v6.11 We have few fixes to drivers. The most important here is a fix for iwlwifi which caused major slowdowns for several users. * tag 'wireless-2024-08-14' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless: wifi: iwlwifi: correctly lookup DMA address in SG table wifi: mt76: mt7921: fix NULL pointer access in mt7921_ipv6_addr_change wifi: brcmfmac: cfg80211: Handle SSID based pmksa deletion wifi: rtlwifi: rtl8192du: Initialise value32 in _rtl92du_init_queue_reserved_page wifi: ath12k: use 128 bytes aligned iova in transmit path for WCN7850 ==================== Link: https://patch.msgid.link/[email protected] Signed-off-by: Jakub Kicinski <[email protected]>
2024-08-14cxgb4: Avoid -Wflex-array-member-not-at-end warningGustavo A. R. Silva1-1/+1
-Wflex-array-member-not-at-end was introduced in GCC-14, and we are getting ready to enable it, globally. Change the type of the middle struct member currently causing trouble from `struct tc_u32_sel` to `struct tc_u32_sel_hdr`. Fix the following warning: drivers/net/ethernet/chelsio/cxgb4/cxgb4_tc_u32_parse.h:245:27: warning: structure containing a flexible array member is not at the end of another structure [-Wflex-array-member-not-at-end] Signed-off-by: Gustavo A. R. Silva <[email protected]> Link: https://patch.msgid.link/97388e8a7990975aa56cf0ada211764c735c3432.1723586870.git.gustavoars@kernel.org Signed-off-by: Jakub Kicinski <[email protected]>
2024-08-14bnxt_en: avoid truncation of per rx run debugfs filenameSimon Horman1-2/+2
Although it seems unlikely in practice - there would need to be rx ring indexes greater than 10^10 - it is theoretically possible for the filename of per rx ring debugfs files to be truncated. This is because although a 16 byte buffer is provided, the length of the filename is restricted to 10 bytes. Remove this restriction and allow the entire buffer to be used. Also reduce the buffer to 12 bytes, which is sufficient. Given that the range of rx ring indexes likely much smaller than the maximum range of a 32-bit signed integer, a smaller buffer could be used, with some further changes. But this change seems simple, robust, and has minimal stack overhead. Flagged by gcc-14: .../bnxt_debugfs.c: In function 'bnxt_debug_dev_init': drivers/net/ethernet/broadcom/bnxt/bnxt_debugfs.c:69:30: warning: '%d' directive output may be truncated writing between 1 and 11 bytes into a region of size 10 [-Wformat-truncation=] 69 | snprintf(qname, 10, "%d", ring_idx); | ^~ In function 'debugfs_dim_ring_init', inlined from 'bnxt_debug_dev_init' at .../bnxt_debugfs.c:87:4: .../bnxt_debugfs.c:69:29: note: directive argument in the range [-2147483643, 2147483646] 69 | snprintf(qname, 10, "%d", ring_idx); | ^~~~ .../bnxt_debugfs.c:69:9: note: 'snprintf' output between 2 and 12 bytes into a destination of size 10 69 | snprintf(qname, 10, "%d", ring_idx); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Compile tested only Signed-off-by: Simon Horman <[email protected]> Reviewed-by: Michael Chan <[email protected]> Link: https://patch.msgid.link/[email protected] Signed-off-by: Jakub Kicinski <[email protected]>
2024-08-14bnxt_en: Extend maximum length of version string by 1 byteSimon Horman1-1/+1
This corrects an out-by-one error in the maximum length of the package version string. The size argument of snprintf includes space for the trailing '\0' byte, so there is no need to allow extra space for it by reducing the value of the size argument by 1. Found by inspection. Compile tested only. Signed-off-by: Simon Horman <[email protected]> Reviewed-by: Michael Chan <[email protected]> Link: https://patch.msgid.link/[email protected] Signed-off-by: Jakub Kicinski <[email protected]>
2024-08-14soc: qcom: socinfo: add QCS8275/QCS8300 SoC IDJingyi Wang1-0/+2
Add SoC ID for Qualcomm QCS8275/QCS8300. Signed-off-by: Jingyi Wang <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: smp2p: use scoped device node handling to simplify error pathsKrzysztof Kozlowski1-12/+4
Obtain the device node reference with scoped/cleanup.h to reduce error handling and make the code a bit simpler. Signed-off-by: Krzysztof Kozlowski <[email protected]> Link: https://lore.kernel.org/r/20240813-b4-cleanup-h-of-node-put-other-v1-6-cfb67323a95c@linaro.org Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: pbs: use scoped device node handling to simplify error pathsKrzysztof Kozlowski1-10/+6
Obtain the device node reference with scoped/cleanup.h to reduce error handling and make the code a bit simpler. Signed-off-by: Krzysztof Kozlowski <[email protected]> Link: https://lore.kernel.org/r/20240813-b4-cleanup-h-of-node-put-other-v1-5-cfb67323a95c@linaro.org Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: ocmem: use scoped device node handling to simplify error pathsKrzysztof Kozlowski1-5/+2
Obtain the device node reference with scoped/cleanup.h to reduce error handling and make the code a bit simpler. Signed-off-by: Krzysztof Kozlowski <[email protected]> Link: https://lore.kernel.org/r/20240813-b4-cleanup-h-of-node-put-other-v1-4-cfb67323a95c@linaro.org Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: ice: use scoped device node handling to simplify error pathsKrzysztof Kozlowski1-9/+5
Obtain the device node reference with scoped/cleanup.h to reduce error handling and make the code a bit simpler. Signed-off-by: Krzysztof Kozlowski <[email protected]> Link: https://lore.kernel.org/r/20240813-b4-cleanup-h-of-node-put-other-v1-3-cfb67323a95c@linaro.org Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: aoss: simplify with scoped for each OF child loopKrzysztof Kozlowski1-5/+3
Use scoped for_each_available_child_of_node_scoped() when iterating over device nodes to make code a bit simpler. Signed-off-by: Krzysztof Kozlowski <[email protected]> Link: https://lore.kernel.org/r/20240813-b4-cleanup-h-of-node-put-other-v1-2-cfb67323a95c@linaro.org Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: apr: simplify with scoped for each OF child loopKrzysztof Kozlowski1-4/+1
Use scoped for_each_child_of_node_scoped() when iterating over device nodes to make code a bit simpler. Signed-off-by: Krzysztof Kozlowski <[email protected]> Link: https://lore.kernel.org/r/20240813-b4-cleanup-h-of-node-put-other-v1-1-cfb67323a95c@linaro.org Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: smd-rpm: add qcom,smd-rpm compatibleDmitry Baryshkov1-0/+6
Add the generic qcom,smd-rpm and qcom,glink-smd-rpm compatibles so that there is no need to add further compat strings to the list. Existing strings are intact to keep compatibility with existing DTS. Signed-off-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14Revert "soc: qcom: smd-rpm: Match rpmsg channel instead of compatible"Dmitry Baryshkov1-9/+26
The rpm_requests device nodes have the compatible node. As such the rpmsg core uses OF modalias instead of a native rpmsg modalias. Thus if smd-rpm is built as a module, it doesn't get autoloaded for the device. Revert the commit bcabe1e09135 ("soc: qcom: smd-rpm: Match rpmsg channel instead of compatible") Fixes: bcabe1e09135 ("soc: qcom: smd-rpm: Match rpmsg channel instead of compatible") Cc: [email protected] Signed-off-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14firmware: qcom: scm: Add multiple download mode supportMukesh Ojha1-2/+5
Currently, scm driver only supports full dump when download mode is selected. Add support to enable minidump as well as enable it along with fulldump. Signed-off-by: Mukesh Ojha <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14firmware: qcom: scm: Refactor code to support multiple dload modeMukesh Ojha2-19/+52
Currently on Qualcomm SoC, download_mode is enabled if CONFIG_QCOM_SCM_DOWNLOAD_MODE_DEFAULT is selected or passed a boolean value from command line. Refactor the code such that it supports multiple download modes and drop CONFIG_QCOM_SCM_DOWNLOAD_MODE_DEFAULT config instead, give interface to set the download mode from module parameter while being backword compatible at the same time. Signed-off-by: Mukesh Ojha <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: pd_mapper: Add more older platforms without domainsStephan Gerhold1-0/+4
MSM8909, MSM8916 and MSM8939 all do not make use of pd-mapper, add them to the list similar to the other older platforms to avoid the following message in dmesg when booting: "PDM: no support for the platform, userspace daemon might be required." Signed-off-by: Stephan Gerhold <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14soc: qcom: pd_mapper: Add X1E80100Stephan Gerhold1-0/+10
X1E80100 has the same protection domains as SM8550, except that MPSS is missing. Add it to the in-kernel pd-mapper to avoid having to run the daemon in userspace for charging and audio functionality. Signed-off-by: Stephan Gerhold <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14Merge branch 'drivers-fixes-for-6.11' into HEADBjorn Andersson5-15/+25
Merge drivers-fixes-for-6.11 branch into drivers-for-6.12 to avoid the (trivial) merge conflict occuring related to the qcom_pdm_domains update.
2024-08-14firmware: qcom: tzmem: fix virtual-to-physical address conversionBartosz Golaszewski1-11/+21
We currently only correctly convert the virtual address passed by the caller to qcom_tzmem_to_phys() if it corresponds to the base address of the chunk. If the user wants to convert some pointer at an offset relative to that base address, we'll return 0. Let's change the implementation of qcom_tzmem_to_phys(): iterate over the chunks and try to call gen_pool_virt_to_phys() just-in-time instead of trying to call it only once when creating the chunk. Fixes: 84f5a7b67b61 ("firmware: qcom: add a dedicated TrustZone buffer allocator") Reported-by: Johan Hovold <[email protected]> Closes: https://lore.kernel.org/lkml/[email protected]/ Acked-by: Andrew Halaney <[email protected]> Signed-off-by: Bartosz Golaszewski <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14firmware: qcom: scm: Mark get_wq_ctx() as atomic callMurali Nalajala1-1/+1
Currently get_wq_ctx() is wrongly configured as a standard call. When two SMC calls are in sleep and one SMC wakes up, it calls get_wq_ctx() to resume the corresponding sleeping thread. But if get_wq_ctx() is interrupted, goes to sleep and another SMC call is waiting to be allocated a waitq context, it leads to a deadlock. To avoid this get_wq_ctx() must be an atomic call and can't be a standard SMC call. Hence mark get_wq_ctx() as a fast call. Fixes: 6bf325992236 ("firmware: qcom: scm: Add wait-queue handling logic") Cc: [email protected] Signed-off-by: Murali Nalajala <[email protected]> Signed-off-by: Unnathi Chalicheemala <[email protected]> Reviewed-by: Elliot Berman <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14remoteproc: qcom: pas: Add support for SA8775p ADSP, CDSP and GPDSPTengfei Fan1-0/+92
Add support for PIL loading on ADSP, CDSP0, CDSP1, GPDSP0 and GPDSP1 on SA8775p SoCs. Signed-off-by: Tengfei Fan <[email protected]> Co-developed-by: Bartosz Golaszewski <[email protected]> Signed-off-by: Bartosz Golaszewski <[email protected]> Link: https://lore.kernel.org/r/20240805-topic-sa8775p-iot-remoteproc-v4-3-86affdc72c04@linaro.org Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14clk: qcom: ipq9574: Update the alpha PLL type for GPLLsdevi priya1-6/+6
Update PLL offsets to DEFAULT_EVO to configure MDIO to 800MHz. The incorrect clock frequency leads to an incorrect MDIO clock. This, in turn, affects the MDIO hardware configurations as the divider is calculated from the MDIO clock frequency. If the clock frequency is not as expected, the MDIO register fails due to the generation of an incorrect MDIO frequency. This issue is critical as it results in incorrect MDIO configurations and ultimately leads to the MDIO function not working. This results in a complete feature failure affecting all Ethernet PHYs. Specifically, Ethernet will not work on IPQ9574 due to this issue. Currently, the clock frequency is set to CLK_ALPHA_PLL_TYPE_DEFAULT. However, this setting does not yield the expected clock frequency. To rectify this, we need to change this to CLK_ALPHA_PLL_TYPE_DEFAULT_EVO. This modification ensures that the clock frequency aligns with our expectations, thereby resolving the MDIO register failure and ensuring the proper functioning of the Ethernet on IPQ9574. Fixes: d75b82cff488 ("clk: qcom: Add Global Clock Controller driver for IPQ9574") Signed-off-by: devi priya <[email protected]> Signed-off-by: Amandeep Singh <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14remoteproc: qcom: pas: Add SDX75 remoteproc supportNaina Mehta1-0/+1
Add MPSS Peripheral Authentication Service support for SDX75 platform. Signed-off-by: Naina Mehta <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14firmware: qcom: scm: Disable SDI and write no dump to dump modeMukesh Ojha1-4/+2
SDI is enabled for most of the Qualcomm SoCs and as per commit ff4aa3bc9825 ("firmware: qcom_scm: disable SDI if required") it was recommended to disable SDI by mentioning it in device tree to avoid hang during watchdog or during reboot. However, for some cases if download mode tcsr register already configured from boot firmware to collect dumps and if SDI is disabled via means of mentioning it in device tree we could still end up with dump collection. Disabling SDI alone is not completely enough to disable dump mode and we also need to zero out the bits download bits from tcsr register. Current commit now, unconditionally call qcom_scm_set_download_mode() based on download_mode flag, at max if TCSR register is not mentioned or available for a SoC it will fallback to legacy way of setting download mode through command which may be no-ops or return error in case current firmware does not implements QCOM_SCM_INFO_IS_CALL_AVAIL so, at worst it does nothing if it fails. It also does to call SDI disable call if dload mode is disabled, which looks fine to do as intention is to disable dump collection even if system crashes. Fixes: ff4aa3bc9825 ("firmware: qcom_scm: disable SDI if required") Signed-off-by: Mukesh Ojha <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14clk: qcom: clk-alpha-pll: Fix CAL_L_VAL override for LUCID EVO PLLAjit Pandey1-1/+1
In LUCID EVO PLL CAL_L_VAL and L_VAL bitfields are part of single PLL_L_VAL register. Update for L_VAL bitfield values in PLL_L_VAL register using regmap_write() API in __alpha_pll_trion_set_rate callback will override LUCID EVO PLL initial configuration related to PLL_CAL_L_VAL bit fields in PLL_L_VAL register. Observed random PLL lock failures during PLL enable due to such override in PLL calibration value. Use regmap_update_bits() with L_VAL bitfield mask instead of regmap_write() API to update only PLL_L_VAL bitfields in __alpha_pll_trion_set_rate callback. Fixes: 260e36606a03 ("clk: qcom: clk-alpha-pll: add Lucid EVO PLL configuration interfaces") Cc: [email protected] Signed-off-by: Ajit Pandey <[email protected]> Reviewed-by: Dmitry Baryshkov <[email protected]> Acked-by: Vladimir Zapolskiy <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Bjorn Andersson <[email protected]>
2024-08-14clk: visconti: Switch to use kmemdup_array()Andy Shevchenko1-3/+3
Let the kmemdup_array() take care about multiplication and possible overflows. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Stephen Boyd <[email protected]>
2024-08-14clk: mmp: Switch to use kmemdup_array()Andy Shevchenko1-6/+4
Let the kmemdup_array() take care about multiplication and possible overflows. Signed-off-by: Andy Shevchenko <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Stephen Boyd <[email protected]>
2024-08-15i2c: tegra: Do not mark ACPI devices as irq safeBreno Leitao1-2/+2
On ACPI machines, the tegra i2c module encounters an issue due to a mutex being called inside a spinlock. This leads to the following bug: BUG: sleeping function called from invalid context at kernel/locking/mutex.c:585 ... Call trace: __might_sleep __mutex_lock_common mutex_lock_nested acpi_subsys_runtime_resume rpm_resume tegra_i2c_xfer The problem arises because during __pm_runtime_resume(), the spinlock &dev->power.lock is acquired before rpm_resume() is called. Later, rpm_resume() invokes acpi_subsys_runtime_resume(), which relies on mutexes, triggering the error. To address this issue, devices on ACPI are now marked as not IRQ-safe, considering the dependency of acpi_subsys_runtime_resume() on mutexes. Fixes: bd2fdedbf2ba ("i2c: tegra: Add the ACPI support") Cc: <[email protected]> # v5.17+ Co-developed-by: Michael van der Westhuizen <[email protected]> Signed-off-by: Michael van der Westhuizen <[email protected]> Signed-off-by: Breno Leitao <[email protected]> Reviewed-by: Dmitry Osipenko <[email protected]> Reviewed-by: Andy Shevchenko <[email protected]> Signed-off-by: Andi Shyti <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.