summaryrefslogtreecommitdiffstats
path: root/drivers
Commit message (Collapse)AuthorAgeFilesLines
...
| * | | | usb: xhci-hcd: Initialize TRT flag for xHCI >= 1.0Andrey Smirnov2019-02-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Initialize TRT flag for xHCI >= 1.0, not just == 1.0. This is what Linux xHCI driver does. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Check usb_pipein(pipe) only once in xhci_submit_normal()Andrey Smirnov2019-02-221-12/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Simplfy xhci_submit_normal()'s logic a bit by using helper variables and making sure that usb_pipein(pipe) is only called once. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Drop 'dma' field from struct xhci_hcdAndrey Smirnov2019-02-222-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Drop 'dma' field from struct xhci_hcd since it is not really used anywhere. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Simplify TRB initialization codeAndrey Smirnov2019-02-221-50/+78
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Current TRB initialization code in xHCI driver is: a) Somewhat inconsistent in endianness correctness: LE is insured in some places whereas in some places it isn't b) Somewhat inconsistent in how DMA buffer address is being written: some places utilize xhci_write_64, whereas others just assing cmd_trb field directly c) A bit wasetful since in a number of codepaths where TRB is used it is memset to zero first only to have 80+% of its fields changed to something else right after. To fix all of the above introduce xhci_init_event_cmd_trb(), that will initialize all of the fileds to desired values only once as well as taking care of endianness. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Convert xhci_submit_control() to use dma_map_single()Andrey Smirnov2019-02-221-7/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Convert xhci_submit_control() to use dma_(un)map_single(). These functions both allow us to handle potential physical/virtual address differences as well as allowing to drop a number of typecasts. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Convert xhci_submit_normal() to use dma_map_single()Andrey Smirnov2019-02-221-7/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Convert xhci_submit_normal() to use dma_(un)map_single(). These functions both allow us to handle potential physical/virtual address differences as well as allowing to drop a number of typecasts. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Don't try to DMA sync if buffer is NULLAndrey Smirnov2019-02-221-8/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Driver's .submit_control() callback can and will be called with buffer set to NULL (and length set to 0), so we need to make sure that we don't try to DMA sync the buffer in that case. Add appropriate gurads to make sure that doesn't happen. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Always wait for "Response Data" completionAndrey Smirnov2019-02-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Xhci_submit_control() submits TRB_DATA with TRB_IOC flag regardless of the vlaue of req->requesttype, so we shouldn't gate waiting for the event that will result from it with "req->requesttype & USB_DIR_IN". Failing to do this will result in unstable USB performance and will eventually cause the controller to fail completely. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Add support for 64-byte context sizeAndrey Smirnov2019-02-222-81/+188
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | xHCI spec allows for various context to be either 32 or 64 bytes in size, so we can't rely on struct definitions to correctly reflect data layout. Port varios context accessors from Linux kernel and convert the rest of the code to use them in order to be able to support both context sizes. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Drop pointless bitwise orAndrey Smirnov2019-02-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On_hs_hub is a boolean, not a bitfiled, so usage of bitwise or is unnecessary. Replace it with a regular assignement operator. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Make sure to initialize xhci->rings_listAndrey Smirnov2019-02-221-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Make sure to initialize xhci->rings_list, otherwise any use of the driver will result in a crash. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Cast ~XHCI_IRQS to u32Andrey Smirnov2019-02-221-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Cast ~XHCI_IRQS to u32 to avoid getting a warning on 64-bit builds. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Do not zero out DMA coherent memoryAndrey Smirnov2019-02-221-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Memory returned by dma_alloc_coherent() should already be zeroed out, so there's no need to do this explicitly. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | usb: xhci-hcd: Tabify the fileAndrey Smirnov2019-02-221-68/+68
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | Replace occasional all-whiltespace indentation with tabs for consistency. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* | | | Merge branch 'for-next/imx-i2c'Sascha Hauer2019-03-075-186/+452
|\ \ \ \
| * | | | i2c: i.MX: Add early driver for use in PBLSascha Hauer2019-03-043-0/+314
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | i2c is used in SPD EEPROMs and thus needed very early before barebox is up. This patch adds a i.MX i2c driver variant for use in PBL. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: Create header file for register definesSascha Hauer2019-02-272-46/+53
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We'll add a stripped down driver variant for use in PBL. To share the register defines add a header file for them. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: Read clock-frequency from device treeSascha Hauer2019-02-271-6/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We speed up the clock frequency only if probed via platform_data. Additionally honour the "clock-frequency" property. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: remove unused defineSascha Hauer2019-02-271-3/+0
| | | | | | | | | | | | | | | | | | | | Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: stop after read only in last transfer of a messageSascha Hauer2019-02-271-3/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Following the Linux driver the controller should be stopped in the read transfer path only after the last transfer of a message, but not when other transfers of the same message are following. Implement this behaviour for the barebox driver aswell. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: Consolidate status polling loopsSascha Hauer2019-02-271-42/+31
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We poll for different bits in the status register in different functions. Instead of repeating the polling loops each time create a helper function for it. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: remove unnecessary ifdefSascha Hauer2019-02-271-8/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | dev_dbg is a no-op when debugging is disabled, so no need for extra ifdefs. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: Fix FSL_I2C_DFSRR register offsetSascha Hauer2019-02-271-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 4ed5b77 ("i.MX: i2c: Add Vybrid support") introduced regshifts for the register offsets, but missed to adjust the register offset for the FSL_I2C_DFSRR register (which exists only on PowerPC). Fix this. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: remove unnecessary ifdefSascha Hauer2019-02-271-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | No need to put the of_compatible field into ifdefs, remove them. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: remove unnecesary defineSascha Hauer2019-02-271-4/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | No need for a define which is used only once. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: fix variable nameSascha Hauer2019-02-271-14/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | i2c_fsl_write() and i2c_fsl_read() take exactly one i2c message, not multiple ones, hence rename the variable from "msgs" to "msg". Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: consolidate codeSascha Hauer2019-02-271-33/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We have to write to FSL_I2C_I2DR and wait for completion/ack three times in the code. Instead of open coding it each time create a helper function for it. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: Track stopped status in I2CR_MSTA bitSascha Hauer2019-02-271-28/+15
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We can track the stopped status in the I2CR_MSTA bit, no need for an extra variable. Also we can call i2c_fsl_stop() instead of open coding it in i2c_fsl_read(). Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: move disabling of controller out of i2c_fsl_stopSascha Hauer2019-02-271-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Move disabling of the controller out of i2c_fsl_stop(). This makes the function reusable in other places in the next patch. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | i2c: i.MX: Do not call i2c_fsl_bus_busy twiceSascha Hauer2019-02-271-4/+0
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | In i2c_fsl_stop() we call i2c_fsl_bus_busy() a second time when it fails. If it fails once it won't succeed the second time, so drop the second call. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* | | | Merge branch 'for-next/imx'Sascha Hauer2019-03-076-9/+434
|\ \ \ \
| * | | | esdhc-xload: check for PRSSTAT_BREN only after each blockSascha Hauer2019-03-051-9/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The BREN bit tells us a watermark level sized buffer is ready for read. Instead of testing it before each FIFO read we must only check it once and then read a watermark level sized buffer. This is at least necessary on Layerscape, otherwise timeouts occur while reading the buffer. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | esdhc-xload: Add bigendian supportSascha Hauer2019-03-051-3/+10
| | | | | | | | | | | | | | | | | | | | Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | esdhc-xload: Use static inline io wrappersSascha Hauer2019-03-051-27/+33
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We'll need big endian support for Layerscape, so put the io accessors into static inline wrappers. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | esdhc-xload: move some register defines to header fileSascha Hauer2019-03-053-10/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | To make them usable for the PBL driver aswell. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | esdhc-xload: Move to drivers/mciSascha Hauer2019-03-053-0/+391
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The esdhc-xload code will be used by upcoming Layerscape support aswell, so move it from architecture code to drivers/mci/. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | | PCI: imx6: Add code to request/control "pcie_aux" clock for i.MX8MQAndrey Smirnov2019-02-271-0/+16
| | |/ / | |/| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Port of a Linux commit 29d6b80bc36be62ae38ed8ac3f7a426975fe7dfa The PCIe IP block has an additional clock, "pcie_aux", that needs to be controlled by the driver. Add code to support it. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com> Reviewed-by: Lucas Stach <l.stach@pengutronix.de> Cc: Bjorn Helgaas <bhelgaas@google.com> Cc: Fabio Estevam <fabio.estevam@nxp.com> Cc: Chris Healy <cphealy@gmail.com> Cc: Lucas Stach <l.stach@pengutronix.de> Cc: Leonard Crestez <leonard.crestez@nxp.com> Cc: "A.s. Dong" <aisheng.dong@nxp.com> Cc: Richard Zhu <hongxing.zhu@nxp.com> Cc: linux-imx@nxp.com Cc: linux-arm-kernel@lists.infradead.org Cc: linux-kernel@vger.kernel.org Cc: linux-pci@vger.kernel.org Cc: Rob Herring <robh@kernel.org> Cc: devicetree@vger.kernel.org Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* | | | Merge branch 'for-next/efi'Sascha Hauer2019-03-077-5/+76
|\ \ \ \ | |_|_|/ |/| | |
| * | | watchdog: add (U)EFI driverOleksij Rempel2019-02-183-0/+71
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This driver is using SetWatchdogTimer() UEFI interface and was tested on iBASE MI991AF Mini-ITX motherboard. Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | | efi: cleanupsOleksij Rempel2019-02-184-5/+5
| |/ / | | | | | | | | | | | | | | | | | | make local functions static and remove unused code Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* | | i.MX habv3: add missing includes and make locally used function staticMichael Olbrich2019-03-041-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Without this building fails with: drivers/hab/habv3.c:80:12: error: static declaration of 'imx25_hab_get_status' follows non-static declaration In file included from drivers/hab/habv3.c:14: include/hab.h:39:5: note: previous declaration of 'imx25_hab_get_status' was here drivers/hab/habv3.c:87:1: warning: data definition has no type or storage class drivers/hab/habv3.c:87:1: warning: type defaults to 'int' in declaration of 'postmmu_initcall' [-Wimplicit-int] drivers/hab/habv3.c:87:1: warning: parameter names (without types) in function declaration drivers/hab/habv3.c:80:12: warning: 'imx25_hab_get_status' defined but not used [-Wunused-function] Signed-off-by: Michael Olbrich <m.olbrich@pengutronix.de> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
* | | Merge branch 'for-next/at91-pmc-clk'Sascha Hauer2019-02-2724-971/+3143
|\ \ \ | |_|/ |/| |
| * | clk: at91: update to PMC bindingsSam Ravnborg2019-02-2516-970/+567
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Based on kernel 5.0-rc6 update at91 clk support to match the new PMC bindings. Manually added all changes done in the kernel from 4.9-rc3 to 5.0-rc6. New drivers required was added as seperate commits. This includes dt-compat code required to support at91sam5d3, as this is not yet ported to use the new PMC bindings. clk-programmable saw some extra changes - it had never been bulit. It is used only by at91sama5d2 - and barebox has no board support for this cpu (yet). The CONFIG_SOC symbols is used to select the relevant drivers. CONFIG_SOC_SAM9 selects several drivers, and in the future this can be split to keep the image size down. In the kernel CLK_OF_DECLARE_DRIVER() can be used for a two step init. In barebox this is a simple one step init. It was added to have less differences between the kernel and the barebox versions of the drivers. Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | clk: at91: add dt-compat to PMC bindingsAhmad Fatoum2019-02-251-0/+961
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As the sama5d3 based microchip-ksz9477-evb board still uses the old bindings, the dt-compat.c code is ported as well. This can be removed when all in-kernel at91 boards have been ported to the new bindings. Signed-off-by: Ahmad Fatoum <a.fatoum@pengutronix.de> Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | clk: at91: add sama5d4Sam Ravnborg2019-02-251-0/+270
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add sama5d4 as part of supporting the PMC bindings for ARM at91. The file will be wired into the build in a follow-up patch. The file is a copy from kernel 5.0-rc6 modified to build with barebox. Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | clk: at91: add sama5d2Sam Ravnborg2019-02-251-0/+342
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add sama5d2 as part of supporting the PMC bindings for ARM at91. The file will be wired into the build in a follow-up patch. The file is a copy from kernel 5.0-rc6 modified to build with barebox. Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | clk: at91: add at91sam9x5Sam Ravnborg2019-02-251-0/+315
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add at91sam9x5 as part of supporting the PMC bindings for ARM at91. The file will be wired into the build in a follow-up patch. The file is a copy from kernel 5.0-rc6 modified to build with barebox. Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | clk: at91: add at91sam9rlSam Ravnborg2019-02-251-0/+177
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add at91sam9rl as part of supporting the PMC bindings for ARM at91. The file will be wired into the build in a follow-up patch. The file is a copy from kernel 5.0-rc6 modified to build with barebox. Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | clk: at91: add at91sam9260Sam Ravnborg2019-02-251-0/+497
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add at91sam9260 as part of supporting the PMC bindings for ARM at91. The file will be wired into the build in a follow-up patch. The file is a copy from kernel 5.0-rc6 modified to build with barebox. Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
| * | clocksource: Do not mix depends and select for ATMEL_PITSam Ravnborg2019-02-251-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | CLOCKSOURCE_ATMEL_PIT was both set to yes using depends on and using select. Use only select to set this non-visible symbol to 'y' Signed-off-by: Sam Ravnborg <sam@ravnborg.org> Tested-by: Ladislav Michl <ladis@linux-mips.org> Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-25 19:08:55 +0000