Page MenuHomeFreeBSD

manu (Emmanuel Vadot)
User

Projects (9)

User Details

User Since
Apr 24 2016, 4:47 PM (552 w, 2 h)

Recent Activity

Today

manu accepted D61863: sysutils/rpi-firmware: Set 32-bit framebuffer depth on Raspberry Pi 5.

I'm not against changing default depth (32bpp seems better to have by default) but it raises two questions :

  • Why does the firmware sets the default to 16bpp ?
  • We still panic with the default firmware configuration ?
Sun, Nov 22, 11:48 AM · arm64, Contributor Reviewers (ports)

Fri, Nov 20

manu added reviewers for D61829: dwc: rockchip: Do not use a NULL clock in the RMII output mode: arm64, br, mmel.
Fri, Nov 20, 6:25 PM
manu added reviewers for D61830: dwc: Do not stall the transmit of the QoS core when a mapping fails: br, mmel.
Fri, Nov 20, 6:25 PM
manu added reviewers for D61831: dwc: Do not panic when a new RX buffer cannot be mapped: br, mmel.
Fri, Nov 20, 6:25 PM
manu added reviewers for D61832: dwc: Set the speed of the SoC from a task: br, mmel.
Fri, Nov 20, 6:25 PM
manu requested review of D61832: dwc: Set the speed of the SoC from a task.
Fri, Nov 20, 6:25 PM
manu requested review of D61831: dwc: Do not panic when a new RX buffer cannot be mapped.
Fri, Nov 20, 6:24 PM
manu requested review of D61830: dwc: Do not stall the transmit of the QoS core when a mapping fails.
Fri, Nov 20, 6:24 PM
manu requested review of D61829: dwc: rockchip: Do not use a NULL clock in the RMII output mode.
Fri, Nov 20, 6:24 PM
manu closed D61705: lindebugfs: Do not expose a node before its private data is set.
Fri, Nov 20, 2:07 PM
manu committed rGd5cbb6229f53: lindebugfs: Do not expose a node before its private data is set.
Fri, Nov 20, 2:07 PM
manu committed rG200b2faf4819: dwc: Limit the DMA of the QoS core to the address width of the hardware.
Fri, Nov 20, 1:56 PM
manu committed rGd67880fca7ff: dwc: rockchip: Do not enable the same clock twice.
Fri, Nov 20, 1:56 PM
manu closed D61018: dwc: Limit the DMA of the QoS core to the address width of the hardware.
Fri, Nov 20, 1:56 PM
manu committed rGf23351e4ff05: dwc: Program the DMA burst length of the QoS core.
Fri, Nov 20, 1:56 PM
manu closed D61016: dwc: rockchip: Do not enable the same clock twice.
Fri, Nov 20, 1:56 PM
manu closed D60920: dwc: Program the DMA burst length of the QoS core.
Fri, Nov 20, 1:56 PM
manu committed rGcfa09932b9aa: dwc: Do not report a negative number of output errors with the QoS core.
Fri, Nov 20, 1:56 PM
manu committed rG9c3faa7da860: dwc: Configure the AXI request limits and mixed burst of the QoS core.
Fri, Nov 20, 1:56 PM
manu closed D60919: dwc: Do not report a negative number of output errors with the QoS core.
Fri, Nov 20, 1:56 PM
manu committed rG57f6320880ab: dwc: rockchip: Add support for the RK3568 GMAC.
Fri, Nov 20, 1:56 PM
manu closed D60918: dwc: Configure the AXI request limits and mixed burst of the QoS core.
Fri, Nov 20, 1:56 PM
manu closed D61019: rk3568_cru: Allow the GMAC RGMII speed clock to reparent.
Fri, Nov 20, 1:56 PM
manu closed D60917: dwc: rockchip: Add support for the RK3568 GMAC.
Fri, Nov 20, 1:56 PM
manu closed D60916: dwc: Use the MAC address from the device tree.
Fri, Nov 20, 1:56 PM
manu closed D61020: clk: Do not print errors when testing a frequency.
Fri, Nov 20, 1:56 PM
manu committed rG9efa26959d3f: dwc: Use the MAC address from the device tree.
Fri, Nov 20, 1:56 PM
manu committed rG39b33f2632ef: clk: Do not print errors when testing a frequency.
Fri, Nov 20, 1:56 PM
manu committed rG8559cae1ebf2: rk3568_cru: Allow the GMAC RGMII speed clock to reparent.
Fri, Nov 20, 1:56 PM
manu requested changes to D61543: bcm2835_fbd: Skip legacy framebuffer setup on BCM2712.

That's just hidding the real problem IMHO

Fri, Nov 20, 1:50 PM · Contributor Reviews (src), arm64
manu abandoned D61017: dwc: Set the packet length in the TX descriptors of the QoS core.

Thanks for the review.
Looking at the RK3568 TRM (section 20.6.3, TDES3 read format): bits [14:0] are FL/TPL,
the length of the packet when TSE is not set,
but "when DWRR/WFQ algorithm is NOT enabled,
value written into this field is not used when TSE = 0".
The driver uses a single TX queue and does not select a scheduling algorithm,
so the hardware never looks at the value and this change has no effect.
Linux and U-Boot do write it, but there is nothing to gain here, so I am dropping the
commit for now and will bring it back if DWRR/WFQ or TSO gets enabled.

Fri, Nov 20, 10:40 AM
manu updated the diff for D61018: dwc: Limit the DMA of the QoS core to the address width of the hardware.

Rebase on latest code (Rebased on D61016)

Fri, Nov 20, 10:40 AM
manu committed R11:59c67976af50: sysutils/rpi-firmware: Use firmware DTB placement on Raspberry Pi 5 (authored by jaewon8868_gmail.com).
Fri, Nov 20, 8:09 AM
manu committed R11:33c820052675: sysutils/rpi-firmware: Set framebuffer_swap=0, no mmc overlay on RPi 4.
Fri, Nov 20, 8:09 AM
manu committed R11:821cc4b7470f: sysutils/rpi-firmware: Set framebuffer_swap=0, no mmc overlay on RPi 4.
Fri, Nov 20, 7:45 AM
manu committed R11:5eb3c528c969: sysutils/u-boot-rpi*: Remove the BGR pixel order patch.
Fri, Nov 20, 7:45 AM

Thu, Nov 19

manu added a comment to D61766: bcm2835_fbd: Use the pixel order to choose the 32 bpp palette.

Related to https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=300542

Thu, Nov 19, 5:39 PM
manu added reviewers for D61766: bcm2835_fbd: Use the pixel order to choose the 32 bpp palette: brd, darkvincentdude_yahoo.com.
Thu, Nov 19, 5:38 PM
manu requested review of D61766: bcm2835_fbd: Use the pixel order to choose the 32 bpp palette.
Thu, Nov 19, 5:37 PM
manu closed D61541: sdhci_fdt: Add BCM2712 SD card support.

Commited as 48fa466426d90bfd02191448d7b81e0da1acee8a

Thu, Nov 19, 1:54 PM · Contributor Reviews (src), arm64
manu requested changes to D61543: bcm2835_fbd: Skip legacy framebuffer setup on BCM2712.

That's likely not the root of the problem, firmware likely created space for a 16bpp framebuffer but the driver is changing it to 24bpp (See https://cgit.freebsd.org/src/tree/sys/arm/broadcom/bcm2835/bcm2835_fbd.c#n88), that's why it crashes.

Thu, Nov 19, 1:52 PM · Contributor Reviews (src), arm64
manu committed rG828f462240be: sdhci_fdt: Add BCM2712 SD card support (authored by jaewon8868_gmail.com).
Thu, Nov 19, 1:50 PM
manu closed D61542: bcm2835_vcbus: Recognize BCM2712.
Thu, Nov 19, 1:49 PM · Contributor Reviews (src), arm64
manu committed rG48fa466426d9: bcm2835_vcbus: Recognize BCM2712 (authored by jaewon8868_gmail.com).
Thu, Nov 19, 1:49 PM
manu accepted D61541: sdhci_fdt: Add BCM2712 SD card support.
Thu, Nov 19, 1:39 PM · Contributor Reviews (src), arm64
manu accepted D61542: bcm2835_vcbus: Recognize BCM2712.
Thu, Nov 19, 1:38 PM · Contributor Reviews (src), arm64

Wed, Nov 18

manu added a comment to D60130: linuxkpi: track cache-attribute lifetimes for overlapping vmap mappings.

@oleglelchuk_gmail.com politely pinging this, there's a few use cases which this helps improve (i915 firmware, drm-kmod unload/load) so I'd like to help land some version of this change to help unblock those efforts

Wed, Nov 18, 3:47 PM
manu added a reviewer for D61705: lindebugfs: Do not expose a node before its private data is set: linuxkpi.
Wed, Nov 18, 8:30 AM
manu requested review of D61705: lindebugfs: Do not expose a node before its private data is set.
Wed, Nov 18, 8:30 AM
manu accepted D61703: pam_xdg.8: Mechanical polish.
Wed, Nov 18, 7:55 AM
manu committed rGee2f66a74d18: linuxkpi: Release devres when the probe of a PCI driver fails.
Wed, Nov 18, 7:42 AM
manu committed rG812bdff114cf: linuxkpi: Release the "from" device in pci_get_class() and pci_get_base_class().
Wed, Nov 18, 7:42 AM
manu closed D61222: linuxkpi: Release the "from" device in pci_get_class() and pci_get_base_class().
Wed, Nov 18, 7:42 AM
manu committed rGc343c445e2c9: lindebugfs: Free the kcov symlink when the module is unloaded.
Wed, Nov 18, 7:42 AM
manu closed D61221: lindebugfs: Free the kcov symlink when the module is unloaded.
Wed, Nov 18, 7:42 AM
manu closed D61534: sysutils/rpi-firmware: Use firmware DTB placement on Raspberry Pi 5.

Commited as https://cgit.freebsd.org/ports/commit/?id=b06f05c6953f35150ab8eaacac273510ea485a2a

Wed, Nov 18, 7:34 AM · arm64
manu closed D61332: pam_xdg: Add linger option.
Wed, Nov 18, 7:29 AM
manu committed rGecd6631b19d9: pam_xdg: Add linger option (authored by ross_ross-williams.net).
Wed, Nov 18, 7:29 AM
manu committed R11:b06f05c6953f: sysutils/rpi-firmware: Use firmware DTB placement on Raspberry Pi 5 (authored by jaewon8868_gmail.com).
Wed, Nov 18, 7:26 AM
manu accepted D61534: sysutils/rpi-firmware: Use firmware DTB placement on Raspberry Pi 5.
Wed, Nov 18, 7:23 AM · arm64
manu added a reviewer for D61694: cpufreq_dt: Read clock-latency-ns and convert latencies to microseconds: arm64.
Wed, Nov 18, 7:01 AM
manu added a reviewer for D61695: cpufreq_dt: Do not count the OPP table nodes that have no opp-hz: arm64.
Wed, Nov 18, 7:01 AM
manu added a reviewer for D61696: cpufreq_dt: Move the parsing of opp-microvolt in its own function: arm64.
Wed, Nov 18, 7:00 AM
manu added a reviewer for D61697: cpufreq_dt: Add a function to resolve an nvmem cell: arm64.
Wed, Nov 18, 7:00 AM
manu added a reviewer for D61698: cpufreq_dt: Add a hook for the SoC specific hardware version: arm64.
Wed, Nov 18, 7:00 AM
manu added a reviewer for D61699: cpufreq_dt: Honor opp-supported-hw and opp-microvolt-<name>: arm64.
Wed, Nov 18, 7:00 AM
manu added a reviewer for D61700: cpufreq_allwinner: Add the hardware version of the H6: arm64.
Wed, Nov 18, 7:00 AM
manu added a reviewer for D61701: cpufreq_dt: Only register the first CPU of a group sharing the OPP table: arm64.
Wed, Nov 18, 7:00 AM
manu requested review of D61701: cpufreq_dt: Only register the first CPU of a group sharing the OPP table.
Wed, Nov 18, 6:59 AM
manu requested review of D61700: cpufreq_allwinner: Add the hardware version of the H6.
Wed, Nov 18, 6:59 AM
manu requested review of D61699: cpufreq_dt: Honor opp-supported-hw and opp-microvolt-<name>.
Wed, Nov 18, 6:59 AM
manu requested review of D61698: cpufreq_dt: Add a hook for the SoC specific hardware version.
Wed, Nov 18, 6:59 AM
manu requested review of D61697: cpufreq_dt: Add a function to resolve an nvmem cell.
Wed, Nov 18, 6:59 AM
manu requested review of D61696: cpufreq_dt: Move the parsing of opp-microvolt in its own function.
Wed, Nov 18, 6:59 AM
manu requested review of D61695: cpufreq_dt: Do not count the OPP table nodes that have no opp-hz.
Wed, Nov 18, 6:59 AM
manu requested review of D61694: cpufreq_dt: Read clock-latency-ns and convert latencies to microseconds.
Wed, Nov 18, 6:59 AM
manu committed rG40e1cc39ad54: aw_thermal: Add support for the H6.
Wed, Nov 18, 6:49 AM
manu committed rG672ff458d112: aw_thermal: Make the data register and the alarm support per SoC.
Wed, Nov 18, 6:49 AM
manu committed rGab350f6e4a6f: aw_thermal: Add an init function to the SoC configuration.
Wed, Nov 18, 6:49 AM
manu committed rGe7b7f2a8c63e: aw_sid: Add the thermal sensor calibration fuse of the H6.
Wed, Nov 18, 6:49 AM
manu committed rGa7b3eb616cf3: aw_sid: Add the CPU speed grade fuse of the H6 and H616.
Wed, Nov 18, 6:49 AM
manu committed rG135c402505bc: ccu_h6: Add the bus gates of UART1, UART2 and UART3.
Wed, Nov 18, 6:48 AM

Tue, Nov 17

manu committed rG5daad0ba461e: twsi: Wait for the bus to go idle before starting a transfer.
Tue, Nov 17, 7:55 AM

Mon, Nov 16

manu committed R11:3c6c14a953f4: sysutils/Makefile: Fix port name.
Mon, Nov 16, 8:19 AM

Sun, Nov 15

manu requested changes to D60882: [PATCH 7/7] aw_gpio: report pinctrl configuration failures.
Sun, Nov 15, 12:00 PM · arm64
manu committed R11:3862a72b6e8b: sysutils/u-boot-cm3588-nas: new port (authored by freebsd_sysctl.cz).
Sun, Nov 15, 11:51 AM
manu accepted D61539: sysutils/u-boot-bananapi-r3: Add new port.
Sun, Nov 15, 11:49 AM
manu added a comment to D61539: sysutils/u-boot-bananapi-r3: Add new port.
In D61539#1411477, @hayzam_gmail.com wrote:

I really don't like having patches for u-boot, where those submitted upstream ?

It is upstream: it's a backport of 3ffad26f17cb ("clk: mediatek: fix using clocks before relocation"), which is in v2027.01-rc1 and only missed 2026.10. Without it the board resets in a loop after BL31. It only applies to this port, and I'll drop it when u-boot-master moves to 2027.01. Or I can hold this port until then if you prefer.

Sun, Nov 15, 11:42 AM
manu requested changes to D57732: arm64: mediatek: add initial MT7622/Banana Pi R64 support.

There is a lot of changes in this review and already a lot of comments, I suggest splitting this into :

Sun, Nov 15, 11:34 AM · arm64
manu requested changes to D61534: sysutils/rpi-firmware: Use firmware DTB placement on Raspberry Pi 5.

So, I haven't followed RPI5 stuff for some time but what's the current behavior without this ? I'm pretty sure that people are using them.
Does that mean that we are currently using DTB loaded by u-boot and you want to switch to the DTB loaded by the firmware ? (like other PI iirc).
I'm not against the change but I'd like to understand more.
Thanks.

Sun, Nov 15, 11:32 AM · arm64
manu accepted D61535: release: Include Raspberry Pi 5 DTB and D0 overlay.
Sun, Nov 15, 11:29 AM · arm64
manu accepted D61540: sysutils/u-boot-bananapi-r3-emmc: Add new port.

As soon as https://reviews.freebsd.org/D61539 is accepted this one can go in.

Sun, Nov 15, 11:29 AM
manu requested changes to D61539: sysutils/u-boot-bananapi-r3: Add new port.

I really don't like having patches for u-boot, where those submitted upstream ?

Sun, Nov 15, 11:28 AM
manu committed rGe23c13a0c967: pci_dw: Honor the max-link-speed property.
Sun, Nov 15, 8:19 AM
manu closed D61014: pci_dw: Honor the max-link-speed property.
Sun, Nov 15, 8:19 AM
manu committed rG495c899ab02b: mcommphy: Set the 10/100 TX delay of the YT8511.
Sun, Nov 15, 8:18 AM
manu closed D61015: mcommphy: Set the 10/100 TX delay of the YT8511.
Sun, Nov 15, 8:18 AM

Sat, Nov 14

manu committed rG512342998935: awg: Stop the controller and power down the PHY on shutdown.
Sat, Nov 14, 6:53 AM

Fri, Nov 13

manu committed rG64abec2e5862: aw_sid: Add support for the H6.
Fri, Nov 13, 11:18 AM
manu committed rGdeedeb1e4160: awg: Use the MAC address of the FDT.
Fri, Nov 13, 11:18 AM