January 2018 - SeaBIOS - mail.coreboot.org

K-TEK-M275TP-FN-BL-ML USB keyboard support
by Аладышев Константин April 28, 2018

April 28, 2018

Hello! I'm having problems with K-TEK-M275TP-FN-BL-ML keyboard (http://www.key-tek.cn/Productview.asp?id=777) It is USB device with keyboard and touchpad combined. It works in OS correctly, but doesn't work in SeaBIOS. I've tried to debug this issue and found that "usb_kbd_setup" function fails on "(epdesc->wMaxPacketSize != 8)" check, cause "epdesc->wMaxPacketSize" is actually 32. I've modified this "if" statement to pass 32 value also and now function completes successfully. But it didn't help to solve my issue. There are no interrupts from keyboard, "ehci_poll_intr" always returns -1 on check: if (token & QTD_STS_ACTIVE) { // No intrs found. return -1; } I don't know what to do now. What should I check next? Does SeaBIOS support USB keyboard+touchpad devices at all?

2 5

[PATCH 0/4] RFC: add CRB TPM device
by marcandre.lureau＠redhat.com Feb. 12, 2018

Feb. 12, 2018

From: Marc-André Lureau <marcandre.lureau(a)redhat.com> Hi, The following series implements a limited TPM CRB driver. The TIS device with a TPM 2.0 seems to be ignored by Windows 10, so I implemented a simple CRB device that I will send shortly on the qemu-devel. With the CRB device, Windows 10 correctly recognized and exchange with a TPM 2.0. As long as the device isn't in qemu, I suppose this series should remain RFC. Feedback welcome! Marc-André Lureau (4): x86: add readq() tpm: generalize init_timeout() tpm: use get_tpm_version() callback WIP: add TPM CRB device support src/hw/tpm_drivers.c | 226 ++++++++++++++++++++++++++++++++++++++++++++++++--- src/hw/tpm_drivers.h | 26 ++++++ src/x86.h | 5 ++ 3 files changed, 245 insertions(+), 12 deletions(-) -- 2.14.1.146.gd35faa819

5 12

Saving a few bytes across a reboot
by Stefan Berger Feb. 8, 2018

Feb. 8, 2018

Kevin, is it possible to save a few bytes, a pointer, across a reboot? I have tried to do this by allocating a memory chunk in the fsegement and storing the pointer there surrounded by 2 'magic' 32 bit values. When trying to find the magic values on reboot early in handle_post() it doesn't seem to find them anymore. Is there another memory segment where SeaBIOS could store the few bytes and find them again? Regards, Stefan

6 27

[Question] Debug Seabios by GDB with Qemu
by Liu, Jing2 Feb. 8, 2018

Feb. 8, 2018

Hello, I am using Seabios for qemu startup, and interested in the debugging by GDB to see how the bios works. Currently, I do it as https://www.seabios.org/Debugging said but met some problems so that couldn't move on. :( Could someone give some pointers on what I missed or have to do to enable the gdb debugging? Note: Arch: x86_64 PC CONFIG_DEBUG_SERIAL is enabled. CONFIG_RELOCATE_INIT is disabled. 1. After Qemu starts, I run "gdb out/rom32seg.o" in another session, and do "target remote localhost:1234", it warns me that "Selected architecture i386 is not compatible with reported target architecture i386:x86_64". But I didn't find any 64bit rom. 2. I can't set any breakpoints (e.g. maininit), because "Function maininit not defined". So where does seabios put the symbol table? And how to break the seabios functions when qemu starts? Looking forward to your response! Thanks, Jing

2 4

[PATCH] Fixes to the floppy driver, so it works on real (not emulated) hardware
by Nikolay Nikolov Feb. 3, 2018

Feb. 3, 2018

Several fixes to the floppy driver, so it works on a real floppy controller with a real floppy. Tested with Coreboot on a motherboard with an ITE IT8712 Super I/O chip (and its built-in floppy controller) and a 1.44MB floppy. Only one floppy is supported for now and only 1.44MB, but more formats will be added in the future. Signed-off-by: Nikolay Nikolov <nickysn(a)users.sourceforge.net> --- src/hw/floppy.c | 83 +++++++++++++++++++++++++++++++++++++++++++++++---------- 1 file changed, 69 insertions(+), 14 deletions(-) diff --git a/src/hw/floppy.c b/src/hw/floppy.c index f2577c5..674909d 100644 --- a/src/hw/floppy.c +++ b/src/hw/floppy.c @@ -34,6 +34,11 @@ #define FLOPPY_GAPLEN 0x1B #define FLOPPY_FORMAT_GAPLEN 0x6c #define FLOPPY_PIO_TIMEOUT 1000 +#define FLOPPY_IRQ_TIMEOUT 5000 +#define FLOPPY_STARTUP_TIME 8 +#define FLOPPY_MOTOR_HOLD 255 // Magic value, means motor must not be turned off +#define FLOPPY_SPECIFY1 0xAF // step rate 12ms, head unload 240ms +#define FLOPPY_SPECIFY2 0x02 // head load time 4ms, DMA used // New diskette parameter table adding 3 parameters from IBM // Since no provisions are made for multiple drive types, most @@ -191,7 +196,8 @@ static void floppy_disable_controller(void) { dprintf(2, "Floppy_disable_controller\n"); - floppy_dor_write(0x00); + // Clear the reset bit (enter reset state) and clear 'enable IRQ and DMA' + floppy_dor_write(GET_LOW(FloppyDOR) & ~0x0c); } static int @@ -199,8 +205,9 @@ floppy_wait_irq(void) { u8 frs = GET_BDA(floppy_recalibration_status); SET_BDA(floppy_recalibration_status, frs & ~FRS_IRQ); + u32 end = timer_calc(FLOPPY_IRQ_TIMEOUT); for (;;) { - if (!GET_BDA(floppy_motor_counter)) { + if (timer_check(end)) { warn_timeout(); floppy_disable_controller(); return DISK_RET_ETIMEOUT; @@ -226,6 +233,7 @@ floppy_wait_irq(void) #define FC_READ (0xe6 | (8<<8) | (7<<12) | FCF_WAITIRQ) #define FC_WRITE (0xc5 | (8<<8) | (7<<12) | FCF_WAITIRQ) #define FC_FORMAT (0x4d | (5<<8) | (7<<12) | FCF_WAITIRQ) +#define FC_SPECIFY (0x03 | (2<<8) | (0<<12)) // Send the specified command and it's parameters to the floppy controller. static int @@ -302,9 +310,12 @@ static int floppy_enable_controller(void) { dprintf(2, "Floppy_enable_controller\n"); - SET_BDA(floppy_motor_counter, FLOPPY_MOTOR_TICKS); - floppy_dor_write(0x00); - floppy_dor_write(0x0c); + // Clear the reset bit (enter reset state), but set 'enable IRQ and DMA' + floppy_dor_write((GET_LOW(FloppyDOR) & ~0x04) | 0x08); + // Real hardware needs a 4 microsecond delay + udelay(4); + // Set the reset bit (normal operation) and keep 'enable IRQ and DMA' on + floppy_dor_write(GET_LOW(FloppyDOR) | 0x0c); int ret = floppy_wait_irq(); if (ret) return ret; @@ -313,6 +324,30 @@ floppy_enable_controller(void) return floppy_pio(FC_CHECKIRQ, param); } +static void +floppy_turn_on_motor(u8 floppyid) +{ + // prevent the motor from being turned off + SET_BDA(floppy_motor_counter, FLOPPY_MOTOR_HOLD); + + u8 motor_mask = 0x10 << floppyid; + if (GET_LOW(FloppyDOR) & motor_mask) { + // If the motor has been started, there's no need to wait for it again + floppy_dor_write(motor_mask | 0x0c | floppyid); + } else { + floppy_dor_write(motor_mask | 0x0c | floppyid); + + msleep(FLOPPY_STARTUP_TIME * 125); + } +} + +static void +floppy_turn_off_motor_delayed(void) +{ + // reset the disk motor timeout value of INT 08 + SET_BDA(floppy_motor_counter, FLOPPY_MOTOR_TICKS); +} + // Activate a drive and send a command to it. static int floppy_drive_pio(u8 floppyid, int command, u8 *param) @@ -324,11 +359,8 @@ floppy_drive_pio(u8 floppyid, int command, u8 *param) return ret; } - // reset the disk motor timeout value of INT 08 - SET_BDA(floppy_motor_counter, FLOPPY_MOTOR_TICKS); - - // Turn on motor of selected drive, DMA & int enabled, normal operation - floppy_dor_write((floppyid ? 0x20 : 0x10) | 0x0c | floppyid); + // Turn on motor of selected drive and wait for it to get up to speed + floppy_turn_on_motor(floppyid); // Send command. int ret = floppy_pio(command, param); @@ -363,6 +395,15 @@ floppy_drive_recal(u8 floppyid) return DISK_RET_SUCCESS; } +static int +floppy_drive_specify(void) +{ + u8 param[2]; + param[0] = FLOPPY_SPECIFY1; + param[1] = FLOPPY_SPECIFY2; + return floppy_pio(FC_SPECIFY, param); +} + static int floppy_drive_readid(u8 floppyid, u8 data_rate, u8 head) { @@ -433,13 +474,23 @@ floppy_prep(struct drive_s *drive_gf, u8 cylinder) !(GET_BDA(floppy_media_state[floppyid]) & FMS_MEDIA_DRIVE_ESTABLISHED)) { // Recalibrate drive. int ret = floppy_drive_recal(floppyid); - if (ret) + if (ret) { + floppy_turn_off_motor_delayed(); return ret; + } // Sense media. ret = floppy_media_sense(drive_gf); - if (ret) + if (ret) { + floppy_turn_off_motor_delayed(); + return ret; + } + + ret = floppy_drive_specify(); + if (ret) { + floppy_turn_off_motor_delayed(); return ret; + } } // Seek to cylinder if needed. @@ -449,8 +500,10 @@ floppy_prep(struct drive_s *drive_gf, u8 cylinder) param[0] = floppyid; param[1] = cylinder; int ret = floppy_drive_pio(floppyid, FC_SEEK, param); - if (ret) + if (ret) { + floppy_turn_off_motor_delayed(); return ret; + } SET_BDA(floppy_track[floppyid], cylinder); } @@ -489,9 +542,11 @@ floppy_dma_cmd(struct disk_op_s *op, int count, int command, u8 *param) dprintf(1, "floppy error: %02x %02x %02x %02x %02x %02x %02x\n" , param[0], param[1], param[2], param[3] , param[4], param[5], param[6]); + floppy_turn_off_motor_delayed(); return DISK_RET_ECONTROLLER; } + floppy_turn_off_motor_delayed(); return DISK_RET_SUCCESS; } @@ -663,7 +718,7 @@ floppy_tick(void) // time to turn off drive(s)? u8 fcount = GET_BDA(floppy_motor_counter); - if (fcount) { + if (fcount && (fcount != FLOPPY_MOTOR_HOLD)) { fcount--; SET_BDA(floppy_motor_counter, fcount); if (fcount == 0) -- 2.14.3

3 4

coreboot: modifying seabios bootenter list dynamically
by Francesco Lucconi Feb. 1, 2018

Feb. 1, 2018

I'm looking for a simple method that enables a dynamic way to select one of the several options listed into CBFS bootorder file but unfortunatelly I'm finding out that this file is static and it could be edited and configured just on build time. My specific goal is about a dynamic OS bootstrapping or via USB storage device or via pci netcard automatically without pressing any keyboard buttons. (the auto mode is guaranteed with a gpio signal that it is switches between these 2 scenarios.) Is it correct? Do you know if I can use it differently?

2 1

[PATCH] build: allow setting tools path from the environment
by Roger Pau Monne Jan. 31, 2018

Jan. 31, 2018

Allow setting the path to as, ld, objcopy, objdump, strip and python from the environment. This is required for building SeaBIOS on FreeBSD, which will switch the default ld to lld very soon, and lld is not capable of building SeaBIOS at the moment. Building SeaBIOS on FreeBSD after the switch to lld will require setting LD=/path/to/gnu/ld at build time. Signed-off-by: Roger Pau Monné <roger.pau(a)citrix.com> --- Cc: Kevin O'Connor <kevin(a)koconnor.net> Cc: Ed Maste <emaste(a)freebsd.org> --- Makefile | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/Makefile b/Makefile index eb8ad58..0fcb1bf 100644 --- a/Makefile +++ b/Makefile @@ -17,12 +17,12 @@ CROSS_PREFIX= ifneq ($(CROSS_PREFIX),) CC=$(CROSS_PREFIX)gcc endif -AS=$(CROSS_PREFIX)as -LD=$(CROSS_PREFIX)ld -OBJCOPY=$(CROSS_PREFIX)objcopy -OBJDUMP=$(CROSS_PREFIX)objdump -STRIP=$(CROSS_PREFIX)strip -PYTHON=python +AS?=$(CROSS_PREFIX)as +LD?=$(CROSS_PREFIX)ld +OBJCOPY?=$(CROSS_PREFIX)objcopy +OBJDUMP?=$(CROSS_PREFIX)objdump +STRIP?=$(CROSS_PREFIX)strip +PYTHON?=python CPP=cpp IASL:=iasl LD32BIT_FLAG:=-melf_i386 -- 2.15.1

2 1

how to configure tft panel resolution in SeaVGABIOS for Geode LX
by Wolfgang Kamp - datakamp Jan. 29, 2018

Jan. 29, 2018

Hi, could anyone explain what is to implement to support different TFT panel sizes and resolutions other than 640x480 in SeaVGABIOS (geodeVGA) for the Geode LX platform? Regards, Wolfgang Kamp

1 0

[PATCH RFC] x86: use volatile asm for read/write{b, w, l} implementations
by Vitaly Kuznetsov Jan. 22, 2018

Jan. 22, 2018

QEMU/KVM guests running nested on top of Hyper-V fail to boot with virtio-blk-pci disks, the debug log ends with Booting from Hard Disk... call32_smm 0x000edd01 e97a0 handle_smi cmd=b5 smbase=0x000a0000 vp notify fe007000 (2) -- 0x0 vp read fe005000 (1) -> 0x0 handle_smi cmd=b5 smbase=0x000a0000 call32_smm done 0x000edd01 0 Booting from 0000:7c00 call32_smm 0x000edd01 e97a4 handle_smi cmd=b5 smbase=0x000a0000 vp notify fe007000 (2) -- 0x0 In resume (status=0) In 32bit resume Attempting a hard reboot ... I bisected the breakage to the following commit: commit f46739b1a819750c63fb5849844d99cc2ab001e8 Author: Kevin O'Connor <kevin(a)koconnor.net> Date: Tue Feb 2 22:34:27 2016 -0500 virtio: Convert to new PCI BAR helper functions But the commit itself appears to be correct. The problem is in how writew() function compiles into vp_notify(). For example, if we drop 'volatile' qualifier from the current writew() implementation everything starts to work. If we disassemble these two versions (as of f46739b1a) the difference will be: 00000000 <vp_notify>: With 'volatile' (current) Without 'volatile' 0: push %ebx 0: push %ebx 1: mov %eax,%ecx 1: mov %eax,%ecx 3: mov 0x1518(%edx),%eax 3: mov 0x1518(%edx),%eax 9: cmpb $0x0,0x2c(%ecx) 9: cmpb $0x0,0x2c(%ecx) d: je 2f <vp_notify+0x2f> d: je 2e <vp_notify+0x2e> f: mov 0x151c(%edx),%edx f: mov 0x151c(%edx),%edx 15: mov 0x28(%ecx),%ebx 18: imul %edx,%ebx 15: imul 0x28(%ecx),%edx 1b: mov 0x8(%ecx),%edx 19: mov 0x8(%ecx),%ebx 1e: add %ebx,%edx 20: cmpb $0x0,0xe(%ecx) 1c: cmpb $0x0,0xe(%ecx) 24: je 2a <vp_notify+0x2a> 20: je 28 <vp_notify+0x28> 22: add %ebx,%edx 26: out %ax,(%dx) 24: out %ax,(%dx) 28: jmp 48 <vp_notify+0x48> 26: jmp 47 <vp_notify+0x47> 2a: mov %ax,(%edx) 28: mov %ax,(%ebx,%edx,1) 2d: jmp 48 <vp_notify+0x48> 2c: jmp 47 <vp_notify+0x47> 2f: lea 0x20(%ecx),%ebx 2e: lea 0x20(%ecx),%ebx 32: cltd 31: cltd 33: push %edx 32: push %edx 34: push %eax 33: push %eax 35: mov $0x2,%ecx 34: mov $0x2,%ecx 3a: mov $0x10,%edx 39: mov $0x10,%edx 3f: mov %ebx,%eax 3e: mov %ebx,%eax 41: call 42 <vp_notify+0x42> 40: call 41 <vp_notify+0x41> 46: pop %eax 45: pop %eax 47: pop %edx 46: pop %edx 48: pop %ebx 47: pop %ebx 49: ret 48: ret My eyes fail to see an obvious compiler flaw here but probably the mov difference (at '2a' old, '28' new) is to blame. Doing some other subtle changes (e.g. adding local variables to the function) help in some cases too. At this point I got a bit lost with my debug so I looked at how Linux does this stuff and it seems we're not using '*(volatile u16) = ' there. Rewriting write/read{b,w,l} with volatile asm help. Signed-off-by: Vitaly Kuznetsov <vkuznets(a)redhat.com> --- RFC: This is rather an ongoing debug as I'm not able to point finger at the real culprit yet, I'd be grateful for any help and suggestions. In particular, I don't quite understand why nested virtualization makes a difference here. --- src/x86.h | 21 +++++++++------------ 1 file changed, 9 insertions(+), 12 deletions(-) diff --git a/src/x86.h b/src/x86.h index 53378e9..d45122c 100644 --- a/src/x86.h +++ b/src/x86.h @@ -199,30 +199,27 @@ static inline void smp_wmb(void) { } static inline void writel(void *addr, u32 val) { - barrier(); - *(volatile u32 *)addr = val; + asm volatile("movl %0, %1" : : "d"(val), "m"(*(u32 *)addr) : "memory"); } static inline void writew(void *addr, u16 val) { - barrier(); - *(volatile u16 *)addr = val; + asm volatile("movw %0, %1" : : "d"(val), "m"(*(u16 *)addr) : "memory"); } static inline void writeb(void *addr, u8 val) { - barrier(); - *(volatile u8 *)addr = val; + asm volatile("movb %0, %1" : : "d"(val), "m"(*(u8 *)addr) : "memory"); } static inline u32 readl(const void *addr) { - u32 val = *(volatile const u32 *)addr; - barrier(); + u32 val; + asm volatile("movl %1, %0" : "=d"(val) : "m"(*(u32 *)addr) : "memory"); return val; } static inline u16 readw(const void *addr) { - u16 val = *(volatile const u16 *)addr; - barrier(); + u16 val; + asm volatile("movw %1, %0" : "=d"(val) : "m"(*(u16 *)addr) : "memory"); return val; } static inline u8 readb(const void *addr) { - u8 val = *(volatile const u8 *)addr; - barrier(); + u8 val; + asm volatile("movb %1, %0" : "=d"(val) : "m"(*(u8 *)addr) : "memory"); return val; } -- 2.14.3

2 6

[PATCH] paravirt: Allow both Red Hat and Amazon PCI HB subsystem vendor and device ids
by Filippo Sironi Jan. 19, 2018

Jan. 19, 2018

Signed-off-by: Filippo Sironi <sironi(a)amazon.de> --- src/fw/paravirt.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/fw/paravirt.c b/src/fw/paravirt.c index 0770c47b12e7..163b70c6473f 100644 --- a/src/fw/paravirt.c +++ b/src/fw/paravirt.c @@ -83,8 +83,8 @@ static void qemu_detect(void) u16 sv = pci_config_readw(0, PCI_SUBSYSTEM_VENDOR_ID); u16 sd = pci_config_readw(0, PCI_SUBSYSTEM_ID); - if (sv != 0x1af4 || /* Red Hat, Inc */ - sd != 0x1100) /* Qemu virtual machine */ + if (!(sv == 0x1af4 && sd == 0x1100) && /* Red Hat, Inc. QEMU VM */ + !(sv == 0x1d0f && sd == 0x1237)) /* Amazon.com, Inc. i440fx VM */ return; PlatformRunningOn |= PF_QEMU; -- 2.7.4

5 4