Branch: refs/heads/coverity_scan
Home: https://github.com/siemens/jailhouse
Commit: 58052a7a9d1f5904d72b1637282c877172ee69f6
https://github.com/siemens/jailhouse/commit/58052a7a9d1f5904d72b1637282c877172ee69f6
Author: Jan Kiszka <[email protected]>
Date: 2019-07-08 (Mon, 08 Jul 2019)
Changed paths:
M VERSION
Log Message:
-----------
Bump version number
Signed-off-by: Jan Kiszka <[email protected]>
Commit: d54cbbcc7c38e9631978756aac1f9f0b9749a80e
https://github.com/siemens/jailhouse/commit/d54cbbcc7c38e9631978756aac1f9f0b9749a80e
Author: Jan Kiszka <[email protected]>
Date: 2019-07-11 (Thu, 11 Jul 2019)
Changed paths:
M inmates/lib/alloc.c
M inmates/lib/include/inmate_common.h
Log Message:
-----------
inmates: Make heap relocatable
The initial heap needs to be within the initially mapped memory in order
to allocate page tables from it for mapping more. On x86, growing the
heap beyond the first MB is blocked by the comm region that has to be
located in lower memory in order to avoid conflicts with PCI resources.
Therefore, it is needed to relocate the heap after mapping its
additional space. Inmates should be able to do that by setting heap_pos.
So, make this pointer global.
Signed-off-by: Jan Kiszka <[email protected]>
Commit: ede099ad87d8d3144aa807cd28b95d287948ad64
https://github.com/siemens/jailhouse/commit/ede099ad87d8d3144aa807cd28b95d287948ad64
Author: Jan Kiszka <[email protected]>
Date: 2019-07-11 (Thu, 11 Jul 2019)
Changed paths:
M configs/x86/qemu-x86.c
Log Message:
-----------
configs: qemu-x86: Fix ACPI region size
At 0x40000000, some PCI resource region starts. We accidentally
overlapped this so far which was unnoticed because ivshmem uses 64-bit
resources, and no other PCI device was placed in that region. When
changing ivshmem to 32-bit, this will become relevant, though.
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 9f233898917f8c1141132606f2f2c624405d8c81
https://github.com/siemens/jailhouse/commit/9f233898917f8c1141132606f2f2c624405d8c81
Author: Jan Kiszka <[email protected]>
Date: 2019-07-13 (Sat, 13 Jul 2019)
Changed paths:
M hypervisor/arch/arm/Makefile
M hypervisor/arch/arm/include/asm/smc.h
M hypervisor/arch/arm/include/asm/sysregs.h
Log Message:
-----------
arm: Fix build with gcc-8
The inline .arch_extension statements are ignored by gcc-8. We rather
need -march=armv7ve now.
To keep older gcc prior to version 5 happy, leave the inline statements
in place. Can be removed once we require newer gcc for other reasons.
Reported-by: Vitaly Andrianov <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: f6fef92ffabab0772a6da1fd15cf12887a9600fa
https://github.com/siemens/jailhouse/commit/f6fef92ffabab0772a6da1fd15cf12887a9600fa
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-16 (Tue, 16 Jul 2019)
Changed paths:
M pyjailhouse/sysfs_parser.py
M tools/root-cell-config.c.tmpl
Log Message:
-----------
pyjailhouse: sysfs_parser: Add more precise length of some extended caps
I often run into situations where Linux tries to write to some extended
capabilities. E.g., Linux wants to clear Advanced Error Reporting (ERR)
registers when probing for a device.
At the moment, the crash dump of the cell is hard to interprete, as it
remains unclear if the access is inside a capability, or if the device
just accesses some PCI config space (e.g., accesses beyond PCI spec). At
the moment, all extended capabilities have a fixed length of 4, which
supports the confusion. Four bytes only cover the header of the
capability.
This patch calculates the correct length of PCI caps for the
configuration file -- at least of some capabilities. For some extended
caps this is pretty easy, as they have a fixed length.
Nevertheless, other vary in their length. In some cases, it's pretty
easy (VNDR, ACS), in other cases it's not worth it (VC, VC9) due to
their complexity.
Caps that aren't handle still result in a length of 4.
Additionally, switch to a hexadecimal representation of the length of
PCI caps in config files.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: ce87a6afc1de4ffdd10fba9c92ff1c89ec7d34e4
https://github.com/siemens/jailhouse/commit/ce87a6afc1de4ffdd10fba9c92ff1c89ec7d34e4
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M hypervisor/arch/arm/include/asm/bitops.h
M hypervisor/arch/arm64/include/asm/bitops.h
M hypervisor/arch/x86/include/asm/bitops.h
Log Message:
-----------
core: bitops: fix type of clear_bit
I don't see a reason why nr should be unsigned for set_bit, but signed for
clear_bit. This probably got copy&pasted across architectures.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 92c1d1d9948fd0c50d6c5bf4259e1dc9ba0b3d4e
https://github.com/siemens/jailhouse/commit/92c1d1d9948fd0c50d6c5bf4259e1dc9ba0b3d4e
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M hypervisor/arch/x86/svm.c
M hypervisor/arch/x86/vmx.c
Log Message:
-----------
x86: use definitions instead of constants
Those definitions are used during allocating, so let's use them during
deallocation as well.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 90f392def0720107d0f0cb17f4d026cfd1c101b2
https://github.com/siemens/jailhouse/commit/90f392def0720107d0f0cb17f4d026cfd1c101b2
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M hypervisor/arch/x86/include/asm/cell.h
M hypervisor/arch/x86/svm.c
M hypervisor/arch/x86/vmx.c
Log Message:
-----------
x86: use the variable io_bitmap for both
Semantically, SVM and VMX don't differ that much in those regards. Let's use
the same variable for both.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 53a55842ef7e9f3695649272b37eed8cec89c271
https://github.com/siemens/jailhouse/commit/53a55842ef7e9f3695649272b37eed8cec89c271
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M hypervisor/arch/x86/include/asm/vcpu.h
M hypervisor/arch/x86/svm.c
M hypervisor/arch/x86/vcpu.c
M hypervisor/arch/x86/vmx.c
Log Message:
-----------
x86: consolidate io_bitmap allocation
It's pretty much the same, besides the number of pages that have to be
allocated.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: c416ff7427f11d51a818bdf8683a75f200f8573d
https://github.com/siemens/jailhouse/commit/c416ff7427f11d51a818bdf8683a75f200f8573d
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M hypervisor/arch/x86/svm.c
M hypervisor/arch/x86/vcpu.c
M hypervisor/arch/x86/vmx.c
Log Message:
-----------
x86: consolidate deallocation of IO bitmaps
Same as before - we can now do this in vcpu.c.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 9ad5ca85ea350eb10dd56a1994183039f5ac13f7
https://github.com/siemens/jailhouse/commit/9ad5ca85ea350eb10dd56a1994183039f5ac13f7
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M hypervisor/arch/x86/vcpu.c
Log Message:
-----------
x86: remove superfluous statement
pio_bitmap already is set to the cell's pio bitmap. No need to repeat the
assignment.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: e4cfe90d48731cf30159d9fb8f9e7283756095f1
https://github.com/siemens/jailhouse/commit/e4cfe90d48731cf30159d9fb8f9e7283756095f1
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M hypervisor/arch/x86/i8042.c
M hypervisor/arch/x86/include/asm/cell.h
M hypervisor/arch/x86/vcpu.c
Log Message:
-----------
x86: simplify check if i8042 needs moderation
By adding pio_i8042_allowed boolean flag to arch cell fields.
With this, we don't need to directly access the pio_bitmap in i8042.c. This
does not only simplify (and speed up) the decision, it will also be helpful in
future patches.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: e7674ea7425fc3f968c6d50a2bffb9afae861f9f
https://github.com/siemens/jailhouse/commit/e7674ea7425fc3f968c6d50a2bffb9afae861f9f
Author: Jan Kiszka <[email protected]>
Date: 2019-07-25 (Thu, 25 Jul 2019)
Changed paths:
M driver/main.c
Log Message:
-----------
driver: Account for renaming of lapic_timer_frequency in 5.3
"lapic_timer_frequency heißt jetzt lapic_timer_period, sonst ändert sich
nix."
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 4f27c75d40daa00e7bb10e2c7b8160c6da8d9732
https://github.com/siemens/jailhouse/commit/4f27c75d40daa00e7bb10e2c7b8160c6da8d9732
Author: Jan Kiszka <[email protected]>
Date: 2019-07-26 (Fri, 26 Jul 2019)
Changed paths:
M Kbuild
Log Message:
-----------
kbuild: Correctly step into driver directory
Descending into subdirectories with modules via subdir-y worked up to
5.3-rc1 but was never designed to do so, see
https://patchwork.kernel.org/patch/11059033. We are supposed to use
obj-m here as well.
Don't migrate the other subdirs because that causes warnings such as
cat: /data/jailhouse/hypervisor/modules.order: No such file or directory
due to missing module targets there.
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 91d4db70c6f9443e27475d76d965ee5e487a61e2
https://github.com/siemens/jailhouse/commit/91d4db70c6f9443e27475d76d965ee5e487a61e2
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-26 (Fri, 26 Jul 2019)
Changed paths:
M configs/arm64/amd-seattle-gic-demo.c
M configs/arm64/amd-seattle-uart-demo.c
M configs/arm64/espressobin-gic-demo.c
M configs/arm64/foundation-v8-gic-demo.c
M configs/arm64/foundation-v8-linux-demo.c
M configs/arm64/foundation-v8-uart-demo.c
M configs/arm64/hikey-gic-demo.c
M configs/arm64/imx8mq-gic-demo.c
M configs/arm64/k3-am654-gic-demo.c
M configs/arm64/k3-am654-uart-demo.c
M configs/arm64/macchiatobin-gic-demo.c
M configs/arm64/miriac-sbc-ls1046a-gic-demo.c
M configs/arm64/qemu-arm64-gic-demo.c
M configs/arm64/ultra96-gic-demo.c
M configs/arm64/zynqmp-zcu102-gic-demo.c
Log Message:
-----------
configs: arm64: don't set .pio_bitmap_size
Needles to set them to zero:
a) The compiler will do it for us
b) ARM64 won't use it
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: a8f1af81a9e2858c9cdc62dcd47471dc5a6e70fa
https://github.com/siemens/jailhouse/commit/a8f1af81a9e2858c9cdc62dcd47471dc5a6e70fa
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-26 (Fri, 26 Jul 2019)
Changed paths:
M hypervisor/arch/x86/include/asm/vcpu.h
M hypervisor/arch/x86/svm.c
M hypervisor/arch/x86/vcpu.c
M hypervisor/arch/x86/vmx.c
Log Message:
-----------
x86: drop vcpu_vendor_get_cell_io_bitmap and struct vcpu_io_bitmap
Intel and AMD are now basically the same. No more need for those level
of indirection. Consolidate it and save a lot of lines of code.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: cb4dee3cfb262cb2a184e7c62a56f1cb812e55c4
https://github.com/siemens/jailhouse/commit/cb4dee3cfb262cb2a184e7c62a56f1cb812e55c4
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-26 (Fri, 26 Jul 2019)
Changed paths:
M include/jailhouse/cell-config.h
Log Message:
-----------
cell-config: introduce pio_whitelist structure and helpers
Introduce struct jailhouse_pio_whitelist, and a macro PIO_RANGE that helps to
fill lists in config files.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 654594bc0be786ae1e495abef7b322e9024f8daa
https://github.com/siemens/jailhouse/commit/654594bc0be786ae1e495abef7b322e9024f8daa
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-26 (Fri, 26 Jul 2019)
Changed paths:
M configs/x86/apic-demo.c
M configs/x86/e1000-demo.c
M configs/x86/f2a88xm-hd3.c
M configs/x86/imb-a180.c
M configs/x86/ioapic-demo.c
M configs/x86/ivshmem-demo.c
M configs/x86/linux-x86-demo.c
M configs/x86/pci-demo.c
M configs/x86/qemu-x86.c
M configs/x86/smp-demo.c
M configs/x86/tiny-demo.c
M hypervisor/arch/x86/vcpu.c
M include/jailhouse/cell-config.h
M tools/jailhouse-hardware-check
Log Message:
-----------
x86: pio: use a whitelist instead of a permission bitmap
Whitelist-based permissions align better with the Jailhouse philosophy.
This patch switches the permission bitmap (which basically was (almost) a
duplicate of the final io_bitmap) to a structure that whitelists a port range,
given a base port and a length.
As a side effect, this patch drastically reduces the size of config files:
8404 -> 224 apic-demo.cell
8488 -> 312 e1000-demo.cell
11450 -> 3290 f2a88xm-hd3.cell
11634 -> 3474 imb-a180.cell
8424 -> 252 ioapic-demo.cell
8480 -> 296 ivshmem-demo.cell
8788 -> 608 linux-x86-demo.cell
8488 -> 308 pci-demo.cell
9730 -> 1586 qemu-x86.cell
8392 -> 212 smp-demo.cell
8404 -> 224 tiny-demo.cell
If no whitelist is given, all PIO access will be denied. Additionally, increase
the config file revision header.
Signed-off-by: Ralf Ramsauer <[email protected]>
[Jan: renamed for_each_pio -> for_each_pio_region]
Signed-off-by: Jan Kiszka <[email protected]>
Commit: ce661a1b2ea6bc66d972a3d7850389f87d9556ce
https://github.com/siemens/jailhouse/commit/ce661a1b2ea6bc66d972a3d7850389f87d9556ce
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-26 (Fri, 26 Jul 2019)
Changed paths:
M tools/jailhouse-cell-linux
M tools/root-cell-config.c.tmpl
Log Message:
-----------
tools: update config generator
Obvious what happens here: Align the root cell template.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 6adfa0af9d1772a9bfa5d2d3a54bb5263a3596e0
https://github.com/siemens/jailhouse/commit/6adfa0af9d1772a9bfa5d2d3a54bb5263a3596e0
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-29 (Mon, 29 Jul 2019)
Changed paths:
M tools/jailhouse-cell-linux
Log Message:
-----------
tools: jailhouse-cell-linux: Simplify calculation of setup_data
The magic constant "25" makes things hard to read. We can abandon this
constant if we let python's struct do the whole calculation.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: ded106d6e583775fb8f9fc704bdfa3181351ac95
https://github.com/siemens/jailhouse/commit/ded106d6e583775fb8f9fc704bdfa3181351ac95
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-29 (Mon, 29 Jul 2019)
Changed paths:
M tools/jailhouse-cell-linux
Log Message:
-----------
tools: jailhouse-cell-linux: Add support for pio_regions
We will need to parse pio_regions in the future.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: 5c45b67ebc7dbdeb959a750d52f226fca48d37ae
https://github.com/siemens/jailhouse/commit/5c45b67ebc7dbdeb959a750d52f226fca48d37ae
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-29 (Mon, 29 Jul 2019)
Changed paths:
M inmates/tools/x86/linux-loader.c
M tools/jailhouse-cell-linux
Log Message:
-----------
inmates: x86: linux-loader: Enrich setup_data with flags
Reserve 32 bits in the setup_data for additional flags. Soon,
jailhouse-cell-linux will use them to indicate the availability of
platform UARTs.
We extended the features of the setup_data, but didn't change the
semantics of any other field. So increase the number of the
compatibility version.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Commit: dee9b4544936cd5c18759c06967f0d52fe147b16
https://github.com/siemens/jailhouse/commit/dee9b4544936cd5c18759c06967f0d52fe147b16
Author: Ralf Ramsauer <[email protected]>
Date: 2019-07-29 (Mon, 29 Jul 2019)
Changed paths:
M tools/jailhouse-cell-linux
Log Message:
-----------
tools: jailhouse-cell-linux: fill the flags
We now have some space for additional flags. We use them to tell Linux
which platform UARTs are available. Linux can use these information to
setup only UARTs that are passed through.
Note that the PIO_RANGEs for UARTs have to be one consecutive
PIO_REGION, e.g. PIO_REGION(0x3f8, 8). Split regions, such as
{ PIO_REGION(0x3f8,4), PIO_REGION(0x3fc, 4) } won't be detected.
Signed-off-by: Ralf Ramsauer <[email protected]>
Signed-off-by: Jan Kiszka <[email protected]>
Compare:
https://github.com/siemens/jailhouse/compare/a17c46ce809e...dee9b4544936
--
You received this message because you are subscribed to the Google Groups
"Jailhouse" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/jailhouse-dev/siemens/jailhouse/push/refs/heads/coverity_scan/a17c46-dee9b4%40github.com.
