Skip to content

GitLab

Switch branch/tag
  1. 28 Jan, 2020 1 commit
  3. 20 Jan, 2020 6 commits
  5. 10 Jan, 2020 2 commits
    • Yann Gautier's avatar
      io: change seek offset to signed long long · 70cb0bff
      Yann Gautier authored 
      

IO seek offset can be set to values above UINT32_MAX, this change
changes the seek offset argument from 'ssize_t' to 'signed long long'.
Fixing platform seek functions to match the new interface update.

Change-Id: I25de83b3b7abe5f52a7b0fee36f71e60cac9cfcb
Signed-off-by: default avatarYann Gautier <yann.gautier@st.com>
Signed-off-by: default avatarEtienne Carriere <etienne.carriere@st.com>
Signed-off-by: default avatarLionel Debieve <lionel.debieve@st.com>
      70cb0bff
    • Deepika Bhavnani's avatar
      Unify type of "cpu_idx" across PSCI module. · 5b33ad17
      Deepika Bhavnani authored 
      

NOTE for platform integrators:
   API `plat_psci_stat_get_residency()` third argument
   `last_cpu_idx` is changed from "signed int" to the
   "unsigned int" type.

Issue / Trouble points
1. cpu_idx is used as mix of `unsigned int` and `signed int` in code
with typecasting at some places leading to coverity issues.

2. Underlying platform API's return cpu_idx as `unsigned int`
and comparison is performed with platform specific defines
`PLAFORM_xxx` which is not consistent

Misra Rule 10.4:
The value of a complex expression of integer type may only be cast to
a type that is narrower and of the same signedness as the underlying
type of the expression.

Based on above points, cpu_idx is kept as `unsigned int` to match
the API's and low-level functions and platform defines are updated
where ever required
Signed-off-by: default avatarDeepika Bhavnani <deepika.bhavnani@arm.com>
Change-Id: Ib26fd16e420c35527204b126b9b91e8babcc3a5c
      5b33ad17
  7. 09 Jan, 2020 1 commit
  9. 06 Jan, 2020 1 commit
  11. 03 Jan, 2020 1 commit
    • Vishnu Banavath's avatar
      drivers: add a driver for snoop control unit · c20c0525
      Vishnu Banavath authored 
      

The SCU connects one to four Cortex-A5/Cortex-A9 processors
to the memory system through the AXI interfaces.

The SCU functions are to:
- maintain data cache coherency between the Cortex-A5/Cortex-A9
  processors
- initiate L2 AXI memory accesses
- arbitrate between Cortex-A5/Cortex-A9 processors requesting
  L2 accesses
- manage ACP accesses.

Snoop Control Unit will enable to snoop on other CPUs caches.
This is very important when it comes to synchronizing data between
CPUs. As an example, there is a high chance that data might be
cache'd and other CPUs can't see the change. In such cases,
if snoop control unit is enabled, data is synchoronized immediately
between CPUs and the changes are visible to other CPUs.

This driver provides functionality to enable SCU as well as enabling
user to know the following
- number of CPUs present
- is a particular CPU operating in SMP mode or AMP mode
- data cache size of a particular CPU
- does SCU has ACP port
- is L2CPRESENT

Change-Id: I0d977970154fa60df57caf449200d471f02312a0
Signed-off-by: default avatarVishnu Banavath <vishnu.banavath@arm.com>
      c20c0525
  13. 30 Dec, 2019 1 commit
    • Andre Przywara's avatar
      console: 16550: Prepare for skipping initialisation · cd50ffd2
      Andre Przywara authored 
      

On some platforms the UART might have already been initialised, for
instance by firmware running before TF-A or by a separate management
processor. In this case it would not be need to initialise it again
(doing so could create spurious characters). But more importantly this
saves us from knowing the right baudrate and the right base clock rate
for the UART. This can lead to more robust and versatile firmware builds.

Allow to skip the 16550 UART initialisation and baud rate divisor
programming, by interpreting an input clock rate of "0" to signify this
case. This will just skip the call to console_16550_core_init, but still
will register the console properly.

Users should just pass 0 as the second parameter, the baudrate (third
parameter) will then be ignored as well.

Fix copy & paste typos in comments for the console_16550_register()
function on the way.
Signed-off-by: default avatarAndre Przywara <andre.przywara@arm.com>
Change-Id: I9f8fca5b358f878fac0f31dc411358fd160786ee
      cd50ffd2
  15. 14 Dec, 2019 2 commits
    • Samuel Holland's avatar
      allwinner: Convert AXP803 regulator setup code into a driver · 0bc752c9
      Samuel Holland authored 
      

Previously, the A64/H5 and H6 platforms' PMIC setup code was entirely
independent. However, some H6 boards also need early regulator setup.

Most of the register interface and all of the device tree traversal code
can be reused between the AXP803 and AXP805. The main difference is the
hardware bus interface, so that part is left to the platforms. The
remainder is moved into a driver.

I factored out the bits that were obviously specific to the AXP803;
additional changes for compatibility with other PMICs can be made as
needed.

The only functional change is that rsb_init() now checks the PMIC's chip
ID register against the expected value. This was already being done in
the H6 version of the code.
Signed-off-by: default avatarSamuel Holland <samuel@sholland.org>
Change-Id: Icdcf9edd6565f78cccc503922405129ac27e08a2
      0bc752c9
    • Samuel Holland's avatar
      drivers: allwinner: axp: Add AXP805 support · f6d9c4ca
      Samuel Holland authored 
      

This adds the new regulator list, as well as changes to make the switch
(equivalent to DC1SW on the AXP803) work on both PMICs.
Signed-off-by: default avatarSamuel Holland <samuel@sholland.org>
Change-Id: I9a1eac8ddfc54b27096c10a8eebdd51aaf9b8311
      f6d9c4ca
  17. 11 Dec, 2019 3 commits
  19. 06 Dec, 2019 1 commit
    • Bence Szépkúti's avatar
      libc: Consolidate unified definitions · b382ac68
      Bence Szépkúti authored 
      

As supporting architectures aside from AArch32 and AArch64 is not a
concern, keeping identical definitions in two places for a large part
of the libc seems counterproductive

The int128 types were left un-unified as __int128 is not supported by
gcc on AArch32
Signed-off-by: default avatarBence Szépkúti <bence.szepkuti@arm.com>
Change-Id: Idf08e6fab7e4680d9da62d3c57266ea2d80472cf
      b382ac68
  21. 19 Nov, 2019 3 commits
    • Justin Chadwell's avatar
      Enable -Wlogical-op always · d7b4cd41
      Justin Chadwell authored 
      

-Wlogical-op prevents common errors with using numerical constants where
a boolean one is expected as well as when the operands of a logical
operator are the same. While these are perfectly valid behavior, they
can be a sign that something is slightly off.

This patch adds this warning to gcc and it's closest equivalent to
clang, while also fixing any warnings that enabling them causes.

Change-Id: Iabadfc1e6ee0c44eef6685a23b0aed8abef8ce89
Signed-off-by: default avatarJustin Chadwell <justin.chadwell@arm.com>
      d7b4cd41
    • Justin Chadwell's avatar
      Enable -Wshadow always · b7f6525d
      Justin Chadwell authored 
      

Variable shadowing is, according to the C standard, permitted and valid
behaviour. However, allowing a local variable to take the same name as a
global one can cause confusion and can make refactoring and bug hunting
more difficult.

This patch moves -Wshadow from WARNING2 into the general warning group
so it is always used. It also fixes all warnings that this introduces
by simply renaming the local variable to a new name

Change-Id: I6b71bdce6580c6e58b5e0b41e4704ab0aa38576e
Signed-off-by: default avatarJustin Chadwell <justin.chadwell@arm.com>
      b7f6525d
    • Max Shvetsov's avatar
      GIC-600: Fix include ordering according to the coding style · ac426351
      Max Shvetsov authored 
      

Change-Id: Ia120bcaacea3a462ab78db13f84ed23493033601
Signed-off-by: default avatarMax Shvetsov <maksims.svecovs@arm.com>
      ac426351
  23. 15 Nov, 2019 1 commit
    • Alexei Fedorov's avatar
      GIC-600: Fix power up sequence · 7a7fbb12
      Alexei Fedorov authored 
      

Arm's GIC-600 features a Power Register (GICR_PWRR),
which needs to be programmed to enable redistributor
operation. Section 3.6.1 in the GIC-600 TRM describes
the power-up and power-down sequence in pseudo code,
which deviates from the current TF-A implementation
in drivers/arm/gic/v3/gic600.c.
For powering on a redistributor, the pseudo code suggests
to loop over the whole sequence (check for transition,
write request bit) instead of just looping over the
ready bit read as TF-A does in gic600_pwr_on().
This patch fixes GIC-600 power up sequence according
to the TRM.

Change-Id: I445c480e96ba356b69a2d8e5308ffe6c0a97f45b
Signed-off-by: default avatarAlexei Fedorov <Alexei.Fedorov@arm.com>
      7a7fbb12
  25. 14 Nov, 2019 1 commit
  27. 11 Nov, 2019 1 commit
  29. 01 Nov, 2019 1 commit
  31. 08 Oct, 2019 1 commit
    • Avinash Mehta's avatar
      Correct UART PL011 initialization calculation · 38d5e150
      Avinash Mehta authored 
      

Currently for Armv7 plaforms the quotient calculated in pl011
uart init code is moved to register r1.

This patch moves the quotient to register r2 as done for other
platforms in the udiv instruction. Value of register r2 is then
used to calculate the values for IBRD and FBRD register

Change-Id: Ie6622f9f0e6d634378b471df5d02823b492c8a24
Signed-off-by: default avatarAvinash Mehta <avinash.mehta@arm.com>
      38d5e150
  33. 03 Oct, 2019 2 commits
  35. 26 Sep, 2019 1 commit
    • Madhukar Pappireddy's avatar
      GICv3: Enable multi socket GIC redistributor frame discovery · ec834925
      Madhukar Pappireddy authored 
      

This patch provides declaration and definition of new GICv3 driver
API: gicv3_rdistif_probe().This function delegates the responsibility
of discovering the corresponding Redistributor base frame to each CPU
itself. It is a modified version of gicv3_rdistif_base_addrs_probe()
and is executed by each CPU in the platform unlike the previous
approach in which only the Primary CPU did the discovery of all the
Redistributor frames for every CPU.

The flush operations as part of gicv3_driver_init() function are
made necessary even for platforms with WARMBOOT_ENABLE_DCACHE_EARLY
because the GICv3 driver data structure contents are accessed by CPU
with D-Cache turned off during power down operations.

Change-Id: I1833e81d3974b32a3e4a3df4766a33d070982268
Signed-off-by: default avatarMadhukar Pappireddy <madhukar.pappireddy@arm.com>
      ec834925
  37. 23 Sep, 2019 1 commit
    • Lionel Debieve's avatar
      stm32mp1: add authentication support for stm32image · 4bdb1a7a
      Lionel Debieve authored 
      

This commit adds authentication binary support for STM32MP1.
It prints the bootrom authentication result if signed
image is used and authenticates the next loaded STM32 images.
It also enables the dynamic translation table support
(PLAT_XLAT_TABLES_DYNAMIC) to use bootrom services.
Signed-off-by: default avatarLionel Debieve <lionel.debieve@st.com>
Change-Id: Iba706519e0dc6b6fae1f3dd498383351f0f75f51
      4bdb1a7a
  39. 20 Sep, 2019 2 commits
  41. 18 Sep, 2019 2 commits
  43. 13 Sep, 2019 3 commits
  45. 12 Sep, 2019 2 commits
    • Justin Chadwell's avatar
      Remove RSA PKCS#1 v1.5 support from cert_tool · 6a415a50
      Justin Chadwell authored 
      Support for PKCS#1 v1.5 was deprecated in SHA 1001202d and fully removed
in SHA fe199e3b

, however, cert_tool is still able to generate
certificates in that form. This patch fully removes the ability for
cert_tool to generate these certificates.

Additionally, this patch also fixes a bug where the issuing certificate
was a RSA and the issued certificate was EcDSA. In this case, the issued
certificate would be signed using PKCS#1 v1.5 instead of RSAPSS per
PKCS#1 v2.1, preventing TF-A from verifying the image signatures. Now
that PKCS#1 v1.5 support is removed, all certificates that are signed
with RSA now use the more modern padding scheme.

Change-Id: Id87d7d915be594a1876a73080528d968e65c4e9a
Signed-off-by: default avatarJustin Chadwell <justin.chadwell@arm.com>
      6a415a50
    • Justin Chadwell's avatar
      Support larger RSA key sizes when using MBEDTLS · aacff749
      Justin Chadwell authored 
      

Previously, TF-A could not support large RSA key sizes as the
configuration options passed to MBEDTLS prevented storing and performing
calculations with the larger, higher-precision numbers required. With
these changes to the arguments passed to MBEDTLS, TF-A now supports
using 3072 (3K) and 4096 (4K) keys in certificates.

Change-Id: Ib73a6773145d2faa25c28d04f9a42e86f2fd555f
Signed-off-by: default avatarJustin Chadwell <justin.chadwell@arm.com>
      aacff749