Re: [PATCH v6 1/8] mmc: sdhci-of-dwcmshc: add common bulk optional clocks support
From: Robin Murphy
Date: Tue Jul 22 2025 - 14:34:06 EST
A bit late for a "review", but Diederik and I have just been
IRC-debugging a crash on RK3568 which by inspection seems to be caused
by this patch:
On 2024-08-05 10:17 am, Chen Wang wrote:
From: Chen Wang <unicorn_wang@xxxxxxxxxxx>
In addition to the required core clock and optional
bus clock, the soc will expand its own clocks, so
the bulk clock mechanism is abstracted.
Note, I call the bulk clocks as "other clocks" due
to the bus clock has been called as "optional".
Signed-off-by: Chen Wang <unicorn_wang@xxxxxxxxxxx>
Tested-by: Drew Fustini <drew@xxxxxxxx> # TH1520
Tested-by: Inochi Amaoto <inochiama@xxxxxxxxxxx> # Duo and Huashan Pi
---
[...]
+static int dwcmshc_get_enable_other_clks(struct device *dev,
+ struct dwcmshc_priv *priv,
+ int num_clks,
+ const char * const clk_ids[])
+{
+ int err;
+
+ if (num_clks > DWCMSHC_MAX_OTHER_CLKS)
+ return -EINVAL;
+
+ for (int i = 0; i < num_clks; i++)
+ priv->other_clks[i].id = clk_ids[i];
+
+ err = devm_clk_bulk_get_optional(dev, num_clks, priv->other_clks);
This leaves a pointer into "priv" in the devres list...
+ if (err) {
+ dev_err(dev, "failed to get clocks %d\n", err);
+ return err;
+ }
+
+ err = clk_bulk_prepare_enable(num_clks, priv->other_clks);
+ if (err)
+ dev_err(dev, "failed to enable clocks %d\n", err);
+
+ priv->num_other_clks = num_clks;
+
+ return err;
+}
+
/*
* If DMA addr spans 128MB boundary, we split the DMA transfer into two
* so that each DMA transfer doesn't exceed the boundary.
[...]
@@ -1280,9 +1300,7 @@ static int dwcmshc_probe(struct platform_device *pdev)
err_clk:
clk_disable_unprepare(pltfm_host->clk);
clk_disable_unprepare(priv->bus_clk);
- if (rk_priv)
- clk_bulk_disable_unprepare(RK35xx_MAX_CLKS,
- rk_priv->rockchip_clks);
+ clk_bulk_disable_unprepare(priv->num_other_clks, priv->other_clks);
free_pltfm:
sdhci_pltfm_free(pdev);
...but upon, say, -EPROBE_DEFER from sdhci_setup_host() because a
regulator isn't ready yet, that "priv" is freed here, so by the time the
devres callbacks eventually run, that "devres->clks" pointer which used
to represent "priv->other_clocks" points to who knows what, and this
sort of thing happens:
[ 12.470827] Unable to handle kernel paging request at virtual address 002df7b378917664
[ 12.472104] Mem abort info:
[ 12.472471] ESR = 0x0000000096000004
[ 12.475991] EC = 0x25: DABT (current EL), IL = 32 bits
[ 12.476657] SET = 0, FnV = 0
[ 12.477146] EA = 0, S1PTW = 0
[ 12.477547] FSC = 0x04: level 0 translation fault
[ 12.478127] Data abort info:
[ 12.478126] rockchip-gpio fdd60000.gpio: probed /pinctrl/gpio@fdd60000
[ 12.478413] ISV = 0, ISS = 0x00000004, ISS2 = 0x00000000
[ 12.479826] CM = 0, WnR = 0, TnD = 0, TagAccess = 0
[ 12.480418] GCS = 0, Overlay = 0, DirtyBit = 0, Xs = 0
[ 12.481282] [002df7b378917664] address between user and kernel address ranges
[ 12.482421] Internal error: Oops: 0000000096000004 [#1] SMP
[ 12.482980] Modules linked in: sdhci_of_dwcmshc drm_dp_aux_bus gpio_rockchip(+) drm_display_helper dw_mmc_rockchip drm_client_lib sdhci_pltfm drm_dma_helper fwnode_mdio sdhci dw_mmc_pltf
m libphy fixed rockchip_dfi drm_kms_helper cqhci pl330(+) phy_rockchip_naneng_combphy dw_wdt phy_rockchip_snps_pcie3 phy_rockchip_inno_usb2 dw_mmc mdio_bus dwc3 ehci_platform ohci_platform
ehci_hcd drm ohci_hcd udc_core io_domain i2c_rk3x usbcore ulpi usb_common
[ 12.486871] CPU: 0 UID: 0 PID: 64 Comm: kworker/u16:3 Not tainted 6.16-rc7-arm64-cknow #1 PREEMPTLAZY Debian 6.16~rc7-1
[ 12.487901] Hardware name: FriendlyElec NanoPi R5S (DT)
[ 12.488412] Workqueue: async async_run_entry_fn
[ 12.488879] pstate: 60400009 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 12.489539] pc : __clk_put+0x2c/0x138
[ 12.489913] lr : __clk_put+0x2c/0x138
[ 12.490281] sp : ffff800080713b10
[ 12.490607] x29: ffff800080713b10 x28: ffff0001f001a120 x27: 0000000000000000
[ 12.491302] x26: ffff0001f98e01a0 x25: 0000000000000000 x24: ffff0001f0f35408
[ 12.491995] x23: ffffa8da199b4b40 x22: ffff800080713bb0 x21: ffff0001f0f35010
[ 12.492689] x20: ffff0001f94aafd0 x19: 0a2df7b378917634 x18: 00000000ffffffff
[ 12.493381] x17: 3d4d455453595342 x16: 555300307075656b x15: ffff0001f4885650
[ 12.494075] x14: 0000000000000000 x13: ffff0001f025b810 x12: 0000000000008000
[ 12.494765] x11: ffffa8da1a73ef98 x10: ffffa8da1a460000 x9 : 0000000000000078
[ 12.495454] x8 : 0000000000000049 x7 : ffffa8da18c2fbe0 x6 : 0000000000000001
[ 12.496145] x5 : 0000000000000004 x4 : 000000006cb6bb63 x3 : 0000000000000000
[ 12.496833] x2 : 0000000000000000 x1 : ffff0001f1365ac0 x0 : 0000000000000001
[ 12.497524] Call trace:
[ 12.497776] __clk_put+0x2c/0x138 (P)
[ 12.498154] clk_put+0x18/0x30
[ 12.498471] clk_bulk_put+0x40/0x68
[ 12.498825] devm_clk_bulk_release+0x24/0x40
[ 12.499248] release_nodes+0x64/0xa0
[ 12.499608] devres_release_all+0x98/0xf8
[ 12.500004] device_unbind_cleanup+0x20/0x70
[ 12.500426] really_probe+0x1e8/0x3a0
[ 12.500793] __driver_probe_device+0x84/0x160
[ 12.501225] driver_probe_device+0x44/0x128
[ 12.501640] __driver_attach_async_helper+0x5c/0x108
[ 12.502125] async_run_entry_fn+0x40/0x180
[ 12.502535] process_one_work+0x23c/0x640
[ 12.502939] worker_thread+0x1b4/0x360
[ 12.503315] kthread+0x150/0x250
[ 12.503646] ret_from_fork+0x10/0x20
[ 12.504015] Code: aa0003f3 b140041f 540006c8 97ffd9c4 (b9403260)
[ 12.504598] ---[ end trace 0000000000000000 ]---
TBH I'm not sure what to do as a straight revert seems impractical by
now, so we hope someone else might have a good idea.
Thanks,
Robin.