Re: [PATCH v1] s390x/tcg: Fix RISBHG

[David Hildenbrand]

On 08.01.21 03:20, Nick Desaulniers wrote:
> On Thu, Jan 7, 2021 at 3:27 PM David Hildenbrand <dhildenb@redhat.com> wrote:
>>
>>
>>> Am 08.01.2021 um 00:21 schrieb Nick Desaulniers <ndesaulniers@google.com>:
>>>
>>> On Thu, Jan 7, 2021 at 3:13 PM David Hildenbrand <david@redhat.com> wrote:
>>>>
>>>> RISBHG is broken and currently hinders clang builds of upstream kernels
>>>> from booting: the kernel crashes early, while decompressing the image.
>>>>
>>>>  [...]
>>>>   Kernel fault: interruption code 0005 ilc:2
>>>>   Kernel random base: 0000000000000000
>>>>   PSW : 0000200180000000 0000000000017a1e
>>>>         R:0 T:0 IO:0 EX:0 Key:0 M:0 W:0 P:0 AS:0 CC:2 PM:0 RI:0 EA:3
>>>>   GPRS: 0000000000000001 0000000c00000000 00000003fffffff4 00000000fffffff0
>>>>         0000000000000000 00000000fffffff4 000000000000000c 00000000fffffff0
>>>>         00000000fffffffc 0000000000000000 00000000fffffff8 00000000008e25a8
>>>>         0000000000000009 0000000000000002 0000000000000008 000000000000bce0
>>>>
>>>> One example of a buggy instruction is:
>>>>
>>>>    17dde:       ec 1e 00 9f 20 5d       risbhg  %r1,%r14,0,159,32
>>>>
>>>> With %r14 = 0x9 and %r1 = 0x7 should result in %r1 = 0x900000007, however,
>>>> results in %r1 = 0.
>>>>
>>>> Let's interpret values of i3/i4 as documented in the PoP and make
>>>> computation of "mask" only based on i3 and i4 and use "pmask" only at the
>>>> very end to make sure wrapping is only applied to the high/low doubleword.
>>>>
>>>> With this patch, I can successfully boot a v5.10 kernel built with
>>>> clang, and gcc builds keep on working.
>>>>
>>>> Fixes: 2d6a869833d9 ("target-s390: Implement RISBG")
>>>> Reported-by: Nick Desaulniers <ndesaulniers@google.com>
>>>> Cc: Guenter Roeck <linux@roeck-us.net>
>>>> Cc: Christian Borntraeger <borntraeger@de.ibm.com>
>>>> Signed-off-by: David Hildenbrand <david@redhat.com>
>>>> ---
>>>>
>>>> This BUG was a nightmare to debug and the code a nightmare to understand.
>>>>
>>>> To make clang/gcc builds boot, the following fix is required as well on
>>>> top of current master: "[PATCH] target/s390x: Fix ALGSI"
>>>> 20210107202135.52379-1-david@redhat.com">https://lkml.kernel.org/r/20210107202135.52379-1-david@redhat.com
>>>
>>> In that case, a huge thank you!!! for this work! ++beers_owed.
>>>
>>
>> :) a kernel build for z13 should work with the (default) „-cpu qemu“ cpu 
>> type.
> 
> Hmm...so I don't think clang can build a Linux kernel image with
> CONFIG_MARCH_Z13=y just yet; just defconfig.  Otherwise looks like
> clang barfs on some of the inline asm constraints.
> 

Ah, right. I overwrote my manual config by a temporary defconfig :)


So, I'm on x86-64 F33.

clang version 11.0.0 (Fedora 11.0.0-2.fc33)
LLVM version 11.0.0

I cannot directly use "LLVM=1" for cross-compilation, as I keep getting
"error: unknown emulation: elf64_s390" from ld.lld and "error: invalid
output format: 'elf64-s390'" from llvm-objcopy. I assume that's fixed in
llvm12?

1. I patch around it (strange, I remember CC= .. used to work, but it no
longer does)

---

index e30cf02da8b8..89c57062ed5d 100644
--- a/Makefile
+++ b/Makefile
@@ -427,13 +427,13 @@ KBUILD_HOSTLDLIBS   := $(HOST_LFS_LIBS) $(HOSTLDLIBS)
 CPP            = $(CC) -E
 ifneq ($(LLVM),)
 CC             = clang
-LD             = ld.lld
-AR             = llvm-ar
-NM             = llvm-nm
-OBJCOPY                = llvm-objcopy
-OBJDUMP                = llvm-objdump
-READELF                = llvm-readelf
-STRIP          = llvm-strip
+LD             = $(CROSS_COMPILE)ld
+AR             = $(CROSS_COMPILE)ar
+NM             = $(CROSS_COMPILE)nm
+OBJCOPY                = $(CROSS_COMPILE)objcopy
+OBJDUMP                = $(CROSS_COMPILE)objdump
+READELF                = $(CROSS_COMPILE)readelf
+STRIP          = $(CROSS_COMPILE)strip
 else
 CC             = $(CROSS_COMPILE)gcc
 LD             = $(CROSS_COMPILE)ld

---

2. Compile using clang


Using latest linux-next (1c925d2030afd354a02c23500386e620e662622b) +
above patch

---

#!/bin/bash
export ARCH=s390;
export CROSS_COMPILE=s390x-linux-gnu-
export LLVM=1
make distclean
make defconfig

# Make F32 initrd boot without inserting modules
./scripts/config -e CONFIG_SCSI_ISCSI_ATTRS
./scripts/config -e CONFIG_ISCSI_TCP

make -j40 > /dev/null

---

3. Run it via QEMU. I boot a full Fedora 32 using the cloud-image +
initrd from Fedora 32 (tried to stick to your cmdline where possible)

./build/qemu-system-s390x \
-m 512M \
-cpu qemu \
-display none \
-nodefaults \
-kernel ../linux-cross/arch/s390/boot/bzImage \
-append "root=/dev/vda1 conmode=sclp console=ttyS0" \
-initrd ../Fedora-Cloud-Base-32-1.6.x86_64-initrd.img \
-hda ../Fedora-Cloud-Base-32-1.6.x86_64-initrd.img \
-serial mon:stdio


KASLR disabled: CPU has no PRNG
[    0.408769] Linux version 5.11.0-rc2-next-20210108-dirty
(dhildenb@desktop) (clang version 11.0.0 (Fedora 11.0.0-2.fc33), GNU ld
version 2.35.1-1.fc33) #1 SMP Fri Jan 8 10:23:01 CET 2021
[    0.410266] setup: Linux is running under KVM in 64-bit mode
[    0.415840] setup: The maximum memory size is 512MB
[    0.417278] cpu: 1 configured CPUs, 0 standby CPUs

...

Fedora 32 (Cloud Edition)
Kernel 5.11.0-rc2-next-20210108-dirty on an s390x (ttysclp0)

atomic-00 login:


> It looks like with your patch applied we get further into the boot!
> I'm not seeing any output with:
> $ /android0/qemu/build/qemu-system-s390x -cpu qemu -append
> 'conmode=sclp console=ttyS0' -display none -initrd
> /<path/to>/boot-utils/images/s390/rootfs.cpio -kernel
> arch/s390/boot/bzImage -m 512m -nodefaults -serial mon:stdio
> 
> (Based on a quick skim through
> https://www.ibm.com/support/knowledgecenter/en/linuxonibm/com.ibm.linux.z.ludd/ludd_r_lmtkernelparameter.html).
> Do I have all of those right?
> 
> If I attach GDB to QEMU running that kernel image, I was able to view
> the print banner once via `lx-dmesg` gdb macro in the kernel, but it
> seems on subsequent runs control flow gets diverted unexpected post
> entry to start_kernel() always to `s390_base_pgm_handler` ...errr..at
> least when I try to single step in GDB.  Tried with linux-5.10.y,
> mainline, and linux-next.
> 
> qemu: 470dd6bd360782f5137f7e3376af6a44658eb1d3 + your patch
> llvm: 106e66f3f555c8f887e82c5f04c3e77bdaf345e8
> linux-5.10.y: d1988041d19dc8b532579bdbb7c4a978391c0011
> linux: 71c061d2443814de15e177489d5cc00a4a253ef3
> linux-next: f87684f6470f5f02bd47d4afb900366e5d2f31b6
> 
> 
> (gdb) hbreak setup_arch
> Hardware assisted breakpoint 1 at 0x142229e: file
> arch/s390/kernel/setup.c, line 1091.
> (gdb) c
> Continuing.
> 
> Program received signal SIGTRAP, Trace/breakpoint trap.
> 0x00000000014222a0 in setup_arch (cmdline_p=0x11d7ed8) at
> arch/s390/kernel/setup.c:1091
> 1091            if (MACHINE_IS_VM)
> (gdb) lx-dmesg
> [    0.376351] Linux version 5.11.0-rc2-00157-ga2885c701c30
> (ndesaulniers@ndesaulniers1.mtv.corp.google.com) (Nick Desaulniers
> clang version 12.0.0 (git@github.com:llvm/llvm-project.git
> e75fec2b238f0e26cfb7645f2208baebe3440d41), GNU ld (GNU Binutils for
> Debian) 2.35.1) #81 SMP Thu Jan 7 17:57:34 PST 2021

So you're using llvm 12. Maybe that makes a difference. Or we have an
issue with our arm64 backend. Or using ld.lld and friends make a
difference. Guess I'd have to custom-compile llvm12 (gah) ... maybe I
can find some rpms somewhere.

-- 
Thanks,

David / dhildenb