linux-amlogic.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
* Regression, bisected: meson-dwmac8b never loads, no ethernet
@ 2021-09-20 23:43 Geraldo Nascimento
       [not found] ` <CAEsQvctayUAp4Cx=TaqjUu90VNZa1itqyfqMNzDrDGfr--aSQg@mail.gmail.com>
  2021-09-21  7:22 ` Arnd Bergmann
  0 siblings, 2 replies; 6+ messages in thread
From: Geraldo Nascimento @ 2021-09-20 23:43 UTC (permalink / raw)
  To: arnd; +Cc: bjorn.andersson, martin.blumenstingl, linux-amlogic

Hello Arnd,

I'm sorry to inform you that commit
a0f480dc6546b9d97027bf7bc8145aad48ddad00 causes a regression in
5.15-rc1 for the meson-dwmac8b ethernet driver.

meson-dwmac8b never loads and libphy never probes stmmac

I've bisected it over the weekend. The results follow.

The bad commit mostly consists of qcom DT changes that shouldn't
trigger anything nasty like missing ethernet for meson (amlogic) but
there's the possibility the regression is caused by stmmac
modifications inside "Merge tag 'v5.14-rc3' into arm64-for-5.15
(2021-08-16 18:02:26 -0500)"

This is just a guess as I haven't been able to pinpoint so far what
precisely causes the regression.

Best,
Geraldo Nascimento

----

Bisecting: a merge base must be tested
[7d2a07b769330c34b4deabeed939325c77a7ec2f] Linux 5.14
Bisecting: 5788 revisions left to test after this (roughly 13 steps)
[1b4f3dfb4792f03b139edf10124fcbeb44e608e6] Merge tag
'usb-serial-5.15-rc1' of
https://git.kernel.org/pub/scm/linux/kernel/git/johan/usb-serial into
usb-next
Bisecting: 2896 revisions left to test after this (roughly 12 steps)
[c793011242d182e5f12800c12dbaf37af80be735] Merge tag 'pinctrl-v5.15-1'
of git://git.kernel.org/pub/scm/linux/kernel/git/linusw/linux-pinctrl
Bisecting: 1446 revisions left to test after this (roughly 11 steps)
[4cdc4cc2ad35f92338497d53d3e8b7876cf2a51d] Merge tag
'asm-generic-5.15' of
git://git.kernel.org/pub/scm/linux/kernel/git/arnd/asm-generic
Bisecting: 703 revisions left to test after this (roughly 10 steps)
[9e5f3ffcf1cb34e7c7beb3f79a96f58536730924] Merge tag
'devicetree-for-5.15' of
git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux
Bisecting: 344 revisions left to test after this (roughly 9 steps)
[a0f480dc6546b9d97027bf7bc8145aad48ddad00] Merge tag
'qcom-arm64-for-5.15' of
git://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux into arm/dt
Bisecting: 197 revisions left to test after this (roughly 8 steps)
[aadf2b3857ad1d09fb28f08058a746f04698f81b] Merge tag
'tegra-for-5.15-arm-dt' of
git://git.kernel.org/pub/scm/linux/kernel/git/tegra/linux into arm/dt
Bisecting: 98 revisions left to test after this (roughly 7 steps)
[f70c6dc013c1842b0ca640926cf74e7b99cec33d] arm64: dts: qcom: pm8150b:
Add DTS node for PMIC VBUS booster
Bisecting: 51 revisions left to test after this (roughly 6 steps)
[6d640913126db109f2e3389173f0fe42413a8183] Merge tag 'imx-dt64-5.15'
of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into
arm/dt
Bisecting: 27 revisions left to test after this (roughly 5 steps)
[9b35ab1e314c3f563f8d6982c0a8f2c8964e4a36] Merge tag 'imx-dt-5.15' of
git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux into
arm/dt
Bisecting: 13 revisions left to test after this (roughly 4 steps)
[77246d45d28f32fee3f92421aaca0e24c97ce6d1] arm64: dts: qcom:
sdm845-mtp: assign DSI clock source parents
Bisecting: 6 revisions left to test after this (roughly 3 steps)
[b05f82b152c905425586f07fb2575011f67a5053] ARM: dts: qcom: msm8974:
Add blsp2_uart7 for bluetooth on sirius
Bisecting: 3 revisions left to test after this (roughly 2 steps)
[f95c4c56d65225a537a2d88735fde7ec4d37641d] ARM: dts: qcom: add ahb
reset to ipq806x-gmac
Bisecting: 1 revision left to test after this (roughly 1 step)
[198b8c8ede36d2135df73fee19506276059a0f8e] Merge tag 'v5.14-rc3' into
arm64-for-5.15
Bisecting: 0 revisions left to test after this (roughly 0 steps)
[4d314179d62b19d0b5ac8de51b491494359c7a19] Merge tag
'qcom-dts-for-5.15' of
git://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux into arm/dt
a0f480dc6546b9d97027bf7bc8145aad48ddad00 is the first bad commit
commit a0f480dc6546b9d97027bf7bc8145aad48ddad00
Merge: 4d314179d62b 198b8c8ede36
Author: Arnd Bergmann <arnd@arndb.de>
Date:   Wed Aug 18 14:37:16 2021 +0200

    Merge tag 'qcom-arm64-for-5.15' of
git://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux into arm/dt

    Qualcomm ARM64 updates for v5.15

    SDM660 and SDM630 was concluded to be similar enough that they should be
    merged, and the derivative SDM636 was added to the bunch. The combined
    platform gained support for GPU, DMA, I2C, IMEM, display, power-domains,
    SDHCI, thermal, USB, interconnects, VADC, WLED and audio remoteproc. The
    Sony Xperia "Ganges" platform was similarly merged with "Nile", got
    cleaned up and gained touchscreen, USB, volume keys and uSD support.

    IPQ6018 gains USB2 and PCIe support and a few minor fixes. IPQ8074
    gains SCM, PRNG and Crypto support and a DT style update of the PCIe
    nodes.

    MSM8916 gains Coresight STM support. The Xiaomi Redmi 2 is introduced,
    with touchscreen, notification LED and IMU support. MSM8996 gains
    support for GPU cooling and v3.0 of the SoC, which is used to introduce
    support for the Sony Xperia X Performance, XZ and XZs phones.

    SC7180 finally gains DisplayPort support and LPASS is updated
    accordingly. A number of fixes are introduced and with the newly
    introduced DRM aux bus in place Trogdor's panel is moved under the eDP
    bridge. SC7280 gained USB, eMMC, SD-card, QFPROM and IPA support, the
    new IDP2 board was added.

    SM6126 (aka Snapdragon 665) was introduced, together with the Sony
    Xperia 10II phone with support for framebuffer, USB, eMMC and volume
    keys.

    SM8150 gained inline crypto support for UFS enabled, CPU opp-tables was
    introduced to scale DDR and L3 frequencies and SPI nodes where added, in
    addition to a number of smaller fixes.

    SM8250 gained a number of minor fixes and had its serial engines wired
    up to use the GENI wrappers' DMA engines.

    SM8350 had wakeup-parent defined for the TLMM gpio node and I2C13 was
    introduced.

    SDM845 display clocks was corrected and Lenovo Yoga C630 got IPA enabled
    and now has working LTE connectivity.

    Additionally a number of minor fixes throughout to correct DT validation
    warnings.

    Lastly v5.14-rc3 is merge in to resolve the merge conflicts caused by
    the USB maintainer deciding to fix a regression in his tree.

    * tag 'qcom-arm64-for-5.15' of
git://git.kernel.org/pub/scm/linux/kernel/git/qcom/linux: (114
commits)
      arm64: dts: qcom: sm8250: assign DSI clock source parents
      arm64: dts: qcom: sdm845-mtp: assign DSI clock source parents
      arm64: dts: qcom: sdm845: assign DSI clock source parents
      arm64: dts: qcom: sc7180: assign DSI clock source parents
      arm64: dts: qcom: sc7280-idp: Add device tree files for IDP2
      dt-bindings: arm: qcom: Document qcom,sc7280-idp2 board
      arm64: dts: qcom: sm8350: fix IPA interconnects
      arm64: dts: qcom: sc7180: define ipa_fw_mem node
      arm64: dts: qcom: sc7280: enable IPA for sc7280-idp
      arm64: dts: qcom: sc7280: add IPA information
      arm64: dts: qcom: sc7180-trogdor: Move panel under the bridge chip
      arm64: dts: qcom: ipq8074: add PRNG node
      arm64: dts: qcom: ipq8074: add crypto nodes
      arm64: dts: qcom: sm8350: add qupv3_id_1/i2c13 nodes
      arm64: dts: qcom: ipq6018: Add pcie support
      arm64: dts: qcom: pm8150b: Add DTS node for PMIC VBUS booster
      arm64: dts: qcom: sm8150: add SPI nodes
      arm64: dts: qcom: msm8916: Enable CoreSight STM component
      arm64: dts: qcom: sc7280: Add qfprom node
      arm64: dts: qcom: sc7280: Fixup the cpufreq node
      ...

    Link: https://lore.kernel.org/r/20210816231223.586597-1-bjorn.andersson@linaro.org
    Signed-off-by: Arnd Bergmann <arnd@arndb.de>

 Documentation/devicetree/bindings/arm/qcom.yaml    |    4 +
 .../devicetree/bindings/thermal/qcom-tsens.yaml    |    1 +
 arch/arm64/boot/dts/qcom/Makefile                  |    9 +
 arch/arm64/boot/dts/qcom/apq8016-sbc.dtsi          |    1 +
 arch/arm64/boot/dts/qcom/apq8096-ifc6640.dts       |    8 +
 arch/arm64/boot/dts/qcom/ipq6018-cp01-c1.dts       |    8 +
 arch/arm64/boot/dts/qcom/ipq6018.dtsi              |  157 ++-
 arch/arm64/boot/dts/qcom/ipq8074-hk01.dts          |    2 +-
 arch/arm64/boot/dts/qcom/ipq8074.dtsi              |   54 +-
 arch/arm64/boot/dts/qcom/msm8916-mtp.dts           |    3 +-
 .../boot/dts/qcom/msm8916-wingtech-wt88047.dts     |  313 +++++
 arch/arm64/boot/dts/qcom/msm8916.dtsi              |   27 +
 arch/arm64/boot/dts/qcom/msm8994.dtsi              |    8 +-
 arch/arm64/boot/dts/qcom/msm8996-mtp.dtsi          |    8 +
 .../qcom/msm8996-pmi8996-sony-xperia-tone-dora.dts |   11 +
 .../msm8996-pmi8996-sony-xperia-tone-kagura.dts    |   11 +
 .../msm8996-pmi8996-sony-xperia-tone-keyaki.dts    |   11 +
 .../dts/qcom/msm8996-sony-xperia-tone-dora.dts     |   27 +
 .../dts/qcom/msm8996-sony-xperia-tone-kagura.dts   |   15 +
 .../dts/qcom/msm8996-sony-xperia-tone-keyaki.dts   |   26 +
 .../boot/dts/qcom/msm8996-sony-xperia-tone.dtsi    |  956 +++++++++++++
 arch/arm64/boot/dts/qcom/msm8996-v3.0.dtsi         |   63 +
 arch/arm64/boot/dts/qcom/msm8996.dtsi              |   44 +-
 arch/arm64/boot/dts/qcom/msm8998.dtsi              |    4 +-
 arch/arm64/boot/dts/qcom/pm660.dtsi                |  133 ++
 arch/arm64/boot/dts/qcom/pm660l.dtsi               |   54 +
 arch/arm64/boot/dts/qcom/pm8004.dtsi               |    2 -
 arch/arm64/boot/dts/qcom/pm8150b.dtsi              |    6 +
 arch/arm64/boot/dts/qcom/pmi8996.dtsi              |   15 +
 arch/arm64/boot/dts/qcom/sa8155p-adp.dts           |   60 +-
 .../arm64/boot/dts/qcom/sc7180-trogdor-coachz.dtsi |    2 +-
 arch/arm64/boot/dts/qcom/sc7180-trogdor.dtsi       |   62 +-
 arch/arm64/boot/dts/qcom/sc7180.dtsi               |  102 +-
 arch/arm64/boot/dts/qcom/sc7280-idp.dts            |  241 +---
 arch/arm64/boot/dts/qcom/sc7280-idp.dtsi           |  341 +++++
 arch/arm64/boot/dts/qcom/sc7280-idp2.dts           |   23 +
 arch/arm64/boot/dts/qcom/sc7280.dtsi               |  419 +++++-
 .../dts/qcom/sdm630-sony-xperia-ganges-kirin.dts   |   14 +-
 .../boot/dts/qcom/sdm630-sony-xperia-ganges.dtsi   |   40 -
 .../dts/qcom/sdm630-sony-xperia-nile-discovery.dts |    1 +
 .../dts/qcom/sdm630-sony-xperia-nile-pioneer.dts   |    1 +
 .../dts/qcom/sdm630-sony-xperia-nile-voyager.dts   |    1 +
 .../boot/dts/qcom/sdm630-sony-xperia-nile.dtsi     |  527 ++++++-
 arch/arm64/boot/dts/qcom/sdm630.dtsi               | 1441 ++++++++++++++++++--
 .../dts/qcom/sdm636-sony-xperia-ganges-mermaid.dts |   14 +-
 arch/arm64/boot/dts/qcom/sdm636.dtsi               |   23 +
 .../arm64/boot/dts/qcom/sdm660-xiaomi-lavender.dts |    2 -
 arch/arm64/boot/dts/qcom/sdm660.dtsi               |  519 +++----
 arch/arm64/boot/dts/qcom/sdm845-mtp.dts            |    3 +
 .../arm64/boot/dts/qcom/sdm845-oneplus-common.dtsi |   11 +
 arch/arm64/boot/dts/qcom/sdm845.dtsi               |   35 +-
 .../boot/dts/qcom/sdm850-lenovo-yoga-c630.dts      |    5 +
 .../dts/qcom/sm6125-sony-xperia-seine-pdx201.dts   |  139 ++
 arch/arm64/boot/dts/qcom/sm6125.dtsi               |  566 ++++++++
 arch/arm64/boot/dts/qcom/sm8150-mtp.dts            |    2 +-
 arch/arm64/boot/dts/qcom/sm8150.dtsi               |  797 ++++++++++-
 arch/arm64/boot/dts/qcom/sm8250.dtsi               |  217 ++-
 arch/arm64/boot/dts/qcom/sm8350-hdk.dts            |    2 +-
 arch/arm64/boot/dts/qcom/sm8350-mtp.dts            |    2 +-
 arch/arm64/boot/dts/qcom/sm8350.dtsi               |   53 +-
 60 files changed, 6754 insertions(+), 892 deletions(-)
 create mode 100644 arch/arm64/boot/dts/qcom/msm8916-wingtech-wt88047.dts
 create mode 100644
arch/arm64/boot/dts/qcom/msm8996-pmi8996-sony-xperia-tone-dora.dts
 create mode 100644
arch/arm64/boot/dts/qcom/msm8996-pmi8996-sony-xperia-tone-kagura.dts
 create mode 100644
arch/arm64/boot/dts/qcom/msm8996-pmi8996-sony-xperia-tone-keyaki.dts
 create mode 100644 arch/arm64/boot/dts/qcom/msm8996-sony-xperia-tone-dora.dts
 create mode 100644 arch/arm64/boot/dts/qcom/msm8996-sony-xperia-tone-kagura.dts
 create mode 100644 arch/arm64/boot/dts/qcom/msm8996-sony-xperia-tone-keyaki.dts
 create mode 100644 arch/arm64/boot/dts/qcom/msm8996-sony-xperia-tone.dtsi
 create mode 100644 arch/arm64/boot/dts/qcom/msm8996-v3.0.dtsi
 create mode 100644 arch/arm64/boot/dts/qcom/pmi8996.dtsi
 create mode 100644 arch/arm64/boot/dts/qcom/sc7280-idp.dtsi
 create mode 100644 arch/arm64/boot/dts/qcom/sc7280-idp2.dts
 delete mode 100644 arch/arm64/boot/dts/qcom/sdm630-sony-xperia-ganges.dtsi
 create mode 100644 arch/arm64/boot/dts/qcom/sdm636.dtsi
 create mode 100644 arch/arm64/boot/dts/qcom/sm6125-sony-xperia-seine-pdx201.dts
 create mode 100644 arch/arm64/boot/dts/qcom/sm6125.dtsi

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Regression, bisected: meson-dwmac8b never loads, no ethernet
       [not found] ` <CAEsQvctayUAp4Cx=TaqjUu90VNZa1itqyfqMNzDrDGfr--aSQg@mail.gmail.com>
@ 2021-09-21  5:31   ` Martin Blumenstingl
  2021-09-21  6:57     ` Neil Armstrong
  0 siblings, 1 reply; 6+ messages in thread
From: Martin Blumenstingl @ 2021-09-21  5:31 UTC (permalink / raw)
  To: Geraldo Nascimento; +Cc: arnd, bjorn.andersson, linux-amlogic

Hi Geraldo,

On Tue, Sep 21, 2021 at 4:09 AM Geraldo Nascimento
<geraldogabriel@gmail.com> wrote:
>
> Sorry to add further noise to this, as there's still 6 release candidates to solve this regression, but the correct name of the driver is meson8b-dwmac.
>
> You see, git bisecting the kernel had my head spinning so much that I apparently switched the proper driver name.
>
> Speaking of which, I swear I ran git bisect every time telling it good when where's good and vice-versa. And was I left with a qcom general device tree commit that also merges qcom 5.14-rc3 into 5.15 because of a regression in USB apparently?
Which board/SoC are you referring to?
I can confirm that dwmac-meson8b still works fine on an Odroid-C1
(Meson8b SoC) with 5.15-rc2
Newer (arm64) SoCs additionally use an MDIO mux driver:
- mdio-mux-mmioreg on GXL, GXM
- mdio-mux-meson-g12a on G12A, G12B, SM1


Best regards,
Martin

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Regression, bisected: meson-dwmac8b never loads, no ethernet
  2021-09-21  5:31   ` Martin Blumenstingl
@ 2021-09-21  6:57     ` Neil Armstrong
  2021-09-21  7:22       ` Geraldo Nascimento
  0 siblings, 1 reply; 6+ messages in thread
From: Neil Armstrong @ 2021-09-21  6:57 UTC (permalink / raw)
  To: Martin Blumenstingl, Geraldo Nascimento
  Cc: arnd, bjorn.andersson, linux-amlogic

Hi Geraldo,

Thanks for the report

On 21/09/2021 07:31, Martin Blumenstingl wrote:
> Hi Geraldo,
> 
> On Tue, Sep 21, 2021 at 4:09 AM Geraldo Nascimento
> <geraldogabriel@gmail.com> wrote:
>>
>> Sorry to add further noise to this, as there's still 6 release candidates to solve this regression, but the correct name of the driver is meson8b-dwmac.
>>
>> You see, git bisecting the kernel had my head spinning so much that I apparently switched the proper driver name.
>>
>> Speaking of which, I swear I ran git bisect every time telling it good when where's good and vice-versa. And was I left with a qcom general device tree commit that also merges qcom 5.14-rc3 into 5.15 because of a regression in USB apparently?
> Which board/SoC are you referring to?
> I can confirm that dwmac-meson8b still works fine on an Odroid-C1
> (Meson8b SoC) with 5.15-rc2
> Newer (arm64) SoCs additionally use an MDIO mux driver:
> - mdio-mux-mmioreg on GXL, GXM
> - mdio-mux-meson-g12a on G12A, G12B, SM1

Indeed it's an issue with mdio muxes, a similar issue had been reported by Marek at https://lore.kernel.org/all/0a2c4106-7f48-2bb5-048e-8c001a7c3fda@samsung.com/
and it concerns commit cf4b94c8530d ("of: property: fw_devlink: Add support for "phy-handle" property")

It may be the same issue here.

Neil

> 
> 
> Best regards,
> Martin
> 
> _______________________________________________
> linux-amlogic mailing list
> linux-amlogic@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-amlogic
> 


_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Regression, bisected: meson-dwmac8b never loads, no ethernet
  2021-09-21  6:57     ` Neil Armstrong
@ 2021-09-21  7:22       ` Geraldo Nascimento
  0 siblings, 0 replies; 6+ messages in thread
From: Geraldo Nascimento @ 2021-09-21  7:22 UTC (permalink / raw)
  To: Neil Armstrong; +Cc: Martin Blumenstingl, arnd, bjorn.andersson, linux-amlogic

Hi Neil and Martin,

On Tue, Sep 21, 2021 at 3:57 AM Neil Armstrong <narmstrong@baylibre.com> wrote:
>
> Hi Geraldo,
>
> Thanks for the report
>
> On 21/09/2021 07:31, Martin Blumenstingl wrote:
> > Hi Geraldo,
> >
> > On Tue, Sep 21, 2021 at 4:09 AM Geraldo Nascimento
> > <geraldogabriel@gmail.com> wrote:
> >>
> >> Sorry to add further noise to this, as there's still 6 release candidates to solve this regression, but the correct name of the driver is meson8b-dwmac.
> >>
> >> You see, git bisecting the kernel had my head spinning so much that I apparently switched the proper driver name.
> >>
> >> Speaking of which, I swear I ran git bisect every time telling it good when where's good and vice-versa. And was I left with a qcom general device tree commit that also merges qcom 5.14-rc3 into 5.15 because of a regression in USB apparently?
> > Which board/SoC are you referring to?
> > I can confirm that dwmac-meson8b still works fine on an Odroid-C1
> > (Meson8b SoC) with 5.15-rc2

At least for me the issue is gone in 5.15-rc2 and present only in 5.15-rc1

My board is is based on the q200 reference design, with 1 GbE external phy.

The only difference between my board and the q200 reference AFAIK is
an Atheros Qualcomm WiFi / BT chip, QCA9377 but this is unrelated.

It's a TV Box branded as H96 Pro+
> > Newer (arm64) SoCs additionally use an MDIO mux driver:
> > - mdio-mux-mmioreg on GXL, GXM
> > - mdio-mux-meson-g12a on G12A, G12B, SM1
>
> Indeed it's an issue with mdio muxes, a similar issue had been reported by Marek at https://lore.kernel.org/all/0a2c4106-7f48-2bb5-048e-8c001a7c3fda@samsung.com/
> and it concerns commit cf4b94c8530d ("of: property: fw_devlink: Add support for "phy-handle" property")
>
> It may be the same issue here.
>

Whatever it was, it seems to be gone at least for me in 5.15-rc2 so it
seems to be low priority unless it reappears.

Thanks,
Geraldo Nascimento

> Neil
>
> >
> >
> > Best regards,
> > Martin
> >
> > _______________________________________________
> > linux-amlogic mailing list
> > linux-amlogic@lists.infradead.org
> > http://lists.infradead.org/mailman/listinfo/linux-amlogic
> >
>

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Regression, bisected: meson-dwmac8b never loads, no ethernet
  2021-09-20 23:43 Regression, bisected: meson-dwmac8b never loads, no ethernet Geraldo Nascimento
       [not found] ` <CAEsQvctayUAp4Cx=TaqjUu90VNZa1itqyfqMNzDrDGfr--aSQg@mail.gmail.com>
@ 2021-09-21  7:22 ` Arnd Bergmann
  2021-09-21  7:26   ` Geraldo Nascimento
  1 sibling, 1 reply; 6+ messages in thread
From: Arnd Bergmann @ 2021-09-21  7:22 UTC (permalink / raw)
  To: Geraldo Nascimento
  Cc: Arnd Bergmann, Bjorn Andersson, Martin Blumenstingl,
	open list:ARM/Amlogic Meson SoC support

On Tue, Sep 21, 2021 at 1:43 AM Geraldo Nascimento
<geraldogabriel@gmail.com> wrote:
>
> Hello Arnd,
>
> I'm sorry to inform you that commit
> a0f480dc6546b9d97027bf7bc8145aad48ddad00 causes a regression in
> 5.15-rc1 for the meson-dwmac8b ethernet driver.
>
> meson-dwmac8b never loads and libphy never probes stmmac
>
> I've bisected it over the weekend. The results follow.
>
> The bad commit mostly consists of qcom DT changes that shouldn't
> trigger anything nasty like missing ethernet for meson (amlogic) but
> there's the possibility the regression is caused by stmmac
> modifications inside "Merge tag 'v5.14-rc3' into arm64-for-5.15
> (2021-08-16 18:02:26 -0500)"
>
> This is just a guess as I haven't been able to pinpoint so far what
> precisely causes the regression.

I'm sure it's not this commit, looking at the contents of the tree
before and after,
it does not touch a single file that is used on your machine.

Sometimes bisections are unreliable. I would suggest you re-test commit
 a0f480dc6546b9 and the one before it, one of them will likely show a different
result from what you had. You can then replay your bisection to find
the actual cause.

       Arnd

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: Regression, bisected: meson-dwmac8b never loads, no ethernet
  2021-09-21  7:22 ` Arnd Bergmann
@ 2021-09-21  7:26   ` Geraldo Nascimento
  0 siblings, 0 replies; 6+ messages in thread
From: Geraldo Nascimento @ 2021-09-21  7:26 UTC (permalink / raw)
  To: Arnd Bergmann
  Cc: Bjorn Andersson, Martin Blumenstingl,
	open list:ARM/Amlogic Meson SoC support

On Tue, Sep 21, 2021 at 4:22 AM Arnd Bergmann <arnd@arndb.de> wrote:
>
> On Tue, Sep 21, 2021 at 1:43 AM Geraldo Nascimento
> <geraldogabriel@gmail.com> wrote:
> >
> > Hello Arnd,
> >
> > I'm sorry to inform you that commit
> > a0f480dc6546b9d97027bf7bc8145aad48ddad00 causes a regression in
> > 5.15-rc1 for the meson-dwmac8b ethernet driver.
> >
> > meson-dwmac8b never loads and libphy never probes stmmac
> >
> > I've bisected it over the weekend. The results follow.
> >
> > The bad commit mostly consists of qcom DT changes that shouldn't
> > trigger anything nasty like missing ethernet for meson (amlogic) but
> > there's the possibility the regression is caused by stmmac
> > modifications inside "Merge tag 'v5.14-rc3' into arm64-for-5.15
> > (2021-08-16 18:02:26 -0500)"
> >
> > This is just a guess as I haven't been able to pinpoint so far what
> > precisely causes the regression.
>
> I'm sure it's not this commit, looking at the contents of the tree
> before and after,
> it does not touch a single file that is used on your machine.
>
> Sometimes bisections are unreliable. I would suggest you re-test commit
>  a0f480dc6546b9 and the one before it, one of them will likely show a different
> result from what you had. You can then replay your bisection to find
> the actual cause.

Thanks, Arnd, I was a bit puzzled to say the least because that commit
seemed totally unrelated like you state yourself.

Like I have mentioned in my other email in reply to Neil and Martin,
issue seems to be gone at least for me in 5.15-rc2.

So unless it reappears I think there's little to do.

Thanks,
Geraldo Nascimento

>
>        Arnd

_______________________________________________
linux-amlogic mailing list
linux-amlogic@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-amlogic

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2021-09-21  7:29 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-09-20 23:43 Regression, bisected: meson-dwmac8b never loads, no ethernet Geraldo Nascimento
     [not found] ` <CAEsQvctayUAp4Cx=TaqjUu90VNZa1itqyfqMNzDrDGfr--aSQg@mail.gmail.com>
2021-09-21  5:31   ` Martin Blumenstingl
2021-09-21  6:57     ` Neil Armstrong
2021-09-21  7:22       ` Geraldo Nascimento
2021-09-21  7:22 ` Arnd Bergmann
2021-09-21  7:26   ` Geraldo Nascimento

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).