Commit Graph
6 Commits
Author SHA1 Message Date
Tim Nordell 43393cf4d1 bootloader: Support for skipping validation upon wake from deep sleep
This saves time when waking up from deep sleep, but potentially decreases
the security of the system.  If the application able to modify itself
(especially areas that are loaded into RAM) in flash while running
without crashing or is modifies the cached bits of information about
what was last booted from the bootloader, this could cause security
issues if the user does a "deep sleep reset" since the full validation
is skipped.

Signed-off-by: Tim Nordell <tim.nordell@nimbelink.com>
2019-08-21 11:44:37 +00:00
Tim Nordell 82984f0539 bootloader: Do not obfuscate RAM if we are not doing image validation
No need to take this step if we are not doing image validation.  The
obfuscation only buys us a tiny bit of "security" anyways since the
main parts of flash are memory mapped, too.  This saves a little bit of
wake-up time when waking up from deep sleep when the
BOOTLOADER_SKIP_VALIDATE_IN_DEEP_SLEEP option is set.

Signed-off-by: Tim Nordell <tim.nordell@nimbelink.com>
2019-08-21 11:44:37 +00:00
Tim Nordell 1e32fa2cf7 bootloader: Remove extraneous newlines from some debug statements
ESP_LOGD(...) provides its own new lines so remove these from these
debug strings.

Signed-off-by: Tim Nordell <tim.nordell@nimbelink.com>
2019-08-21 11:44:37 +00:00
Tim Nordell 7892cf6a03 partition: Fix "encrypted" read/write when encryption is disabled
According to the documentation[1][2] for partitions, setting the encrypted
flag for partitions should be a no-op when system level encryption isn't
enabled.  The current implementation, however, does not actually match
the documentation and it ends up with an unreadable partition via the
partition API if a partition flag is marked as encrypted without
system-level encryption enabled.  (This is because the writes go through
the encryption block, and reads do not go through the encryption block
when this situation occurs causing unreadable data to the application
running.) This fixes up the read-back of the partition table to match
whether or not the partition is currently encrypted under the hood.

This should not affect the bootloader's code for reading/writing encrypted
partitions as the bootloader directly invokes the spi_flash_write*(...)
APIs.

[1] https://docs.espressif.com/projects/esp-idf/en/latest/api-guides/partition-tables.html#flags
[2] https://docs.espressif.com/projects/esp-idf/en/latest/security/flash-encryption.html#encrypted-partition-flag

Closes https://github.com/espressif/esp-idf/pull/3328

Signed-off-by: Tim Nordell <tim.nordell@nimbelink.com>
2019-05-03 15:03:04 +05:30
Tim Nordell c3fa2497b9 nvs: Check if an item is modified before writing out an identical copy
This prevents wear and tear on the flash, and it also is faster in some
cases since the read-out of flash is a cheaper operation than the erasure
of flash.  Some library modules (such as the esp_wifi) write out to NVS
upon every initialization without checking first that the existing value
is the same, and this speeds up initialization of modules that make
these design choices and moves it into a centralized place.

The comparison functions are based on the read-out functions of the same
name, and changes out the memcpy(...) operations for memcmp(...)
operations.

Signed-off-by: Tim Nordell <tim.nordell@nimbelink.com>
2019-04-30 11:39:58 -05:00
Tim Nordell eb5abca23d nvs_flash: Fix typo in README.rst
Merges https://github.com/espressif/esp-idf/pull/3216

This should read "except for the new key-value pair" instead of
"expect for the new key-value pair".

Signed-off-by: Tim Nordell <tim.nordell@nimbelink.com>
2019-03-24 19:56:10 +01:00