linux.git/arch/s390/boot, branch v4.20

s390: avoid vmlinux segments overlap

2018-10-26T08:19:40+00:00

Currently .vmlinux.info section of uncompressed vmlinux elf image is
included into the data segment and load address specified as 0. That
extends data segment to address 0 and makes "text" and "data" segments
overlap.
Program Headers:
  Type           Offset             VirtAddr           PhysAddr
                 FileSiz            MemSiz              Flags  Align
  LOAD           0x0000000000001000 0x0000000000100000 0x0000000000100000
                 0x0000000000ead03c 0x0000000000ead03c  R E    0x1000
  LOAD           0x0000000000eaf000 0x0000000000000000 0x0000000000000000
                 0x0000000001a13400 0x000000000233b520  RWE    0x1000
  NOTE           0x0000000000eae000 0x0000000000fad000 0x0000000000fad000
                 0x000000000000003c 0x000000000000003c         0x4

 Section to Segment mapping:
  Segment Sections...
   00     .text .notes
   01     .rodata __ksymtab __ksymtab_gpl __ksymtab_strings __param
   __modver .data..ro_after_init __ex_table .data __bug_table .init.text
   .exit.text .exit.data .altinstructions .altinstr_replacement
   .nospec_call_table .nospec_return_table .boot.data .init.data
   .data..percpu .bss .vmlinux.info
   02     .notes

Later when vmlinux.bin is produced from vmlinux, .vmlinux.info section
is removed. But elf vmlinux file, even though it is not bootable anymore,
used for debugging and loadable segments overlap should be avoided.

Utilize special ":NONE" phdr specification to avoid adding .vmlinux.info
into loadable data segment. Also set .vmlinux.info section type to INFO,
which allows to get a not-loadable info CONTENTS section.

Since minimal supported version of binutils 2.20 does not have
--dump-section objcopy option, make .vmlinux.info section loadable during
info.bin creation to get actual section contents.

Reported-by: Philipp Rudo 
Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390/decompressor: add missing FORCE to build targets

2018-10-26T08:19:40+00:00

According to Documentation/kbuild/makefiles.txt all build targets
using if_changed should use FORCE as well. Add missing FORCE to make
sure vmlinux decompressor targets are rebuild properly when not just
immediate prerequisites have changed but also when build command differs.

Reviewed-by: Philipp Rudo 
Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390/mem_detect: add missing include

2018-10-10T05:37:22+00:00

Fix this allnoconfig build breakage:

arch/s390/boot/mem_detect.c: In function 'tprot':
arch/s390/boot/mem_detect.c:122:12: error: 'EFAULT' undeclared (first use in this function)

Signed-off-by: Heiko Carstens 
Signed-off-by: Martin Schwidefsky

s390/head: avoid doubling early boot stack size under KASAN

2018-10-09T09:21:32+00:00

Early boot stack uses predefined 4 pages of memory 0x8000-0xC000. This
stack is used to run not instumented decompressor/facilities
verification C code. It doesn't make sense to double its size when
the kernel is built with KASAN support. BOOT_STACK_ORDER is introduced
to avoid that.

Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390/kasan: use noexec and large pages

2018-10-09T09:21:24+00:00

To lower memory footprint and speed up kasan initialisation detect
EDAT availability and use large pages if possible. As we know how
much memory is needed for initialisation, another simplistic large
page allocator is introduced to avoid memory fragmentation.

Since facilities list is retrieved anyhow, detect noexec support and
adjust pages attributes. Handle noexec kernel option to avoid inconsistent
kasan shadow memory pages flags.

Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390/kasan: avoid instrumentation of early C code

2018-10-09T09:21:17+00:00

Instrumented C code cannot run without the kasan shadow area. Exempt
source code files from kasan which are running before / used during
kasan initialization.

Reviewed-by: Martin Schwidefsky 
Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390: move ipl block and cmd line handling to early boot phase

2018-10-09T09:21:14+00:00

To distinguish zfcpdump case and to be able to parse some of the kernel
command line arguments early (e.g. mem=) ipl block retrieval and command
line construction code is moved to the early boot phase.

"memory_end" is set up correctly respecting "mem=" and hsa_size in case
of the zfcpdump.

arch/s390/boot/string.c is introduced to provide string handling and
command line parsing functions to early boot phase code for the compressed
kernel image case.

Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390/mem_detect: replace tprot loop with binary search

2018-10-09T09:21:12+00:00

In a situation when other memory detection methods are not available
(no SCLP and no z/VM diag260), continuous online memory is assumed.
Replacing tprot loop with faster binary search, as only online memory
end has to be found.

Reviewed-by: Heiko Carstens 
Reviewed-by: Martin Schwidefsky 
Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390/mem_detect: use SCLP info for continuous memory detection

2018-10-09T09:21:11+00:00

When neither SCLP storage info, nor z/VM diag260 "storage configuration"
are available assume a continuous online memory of size specified by
SCLP info.

Reviewed-by: Heiko Carstens 
Reviewed-by: Martin Schwidefsky 
Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky

s390/mem_detect: introduce z/VM specific diag260 call

2018-10-09T09:21:10+00:00

In the case when z/VM memory is defined with "define storage config"
command, SCLP storage info is not available. Utilize diag260 "storage
configuration" call, to get information about z/VM specific guest memory
definitions with potential memory holes.

Reviewed-by: Martin Schwidefsky 
Signed-off-by: Vasily Gorbik 
Signed-off-by: Martin Schwidefsky