Conversation
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
| // Shared heap for memory-mapped access | ||
| wasm_shared_heap_t _shared_heap = NULL; | ||
| #ifdef CONFIG_OCRE_SHARED_HEAP_BUF_VIRTUAL | ||
| uint8 preallocated_buf[CONFIG_OCRE_SHARED_HEAP_BUF_SIZE]; |
There was a problem hiding this comment.
Should we allocate and put this in the external ram (storage_malloc) or (user_malloc) instead of making it global/static?
There was a problem hiding this comment.
I am fine with either - but with current solution we would still use system memory (just allocated dynamically) on systems with no PSRAM, will need to cut a bit WAMR's linear memory and allocate/deallocate memory during the runtime init.
I propose the following: we keep it this way as initial commit for this features, and once the shared_memory is available and non-kernel stuff are using the same pull, we will move everything (inc. this) to use that.
|
|
||
| #ifdef CONFIG_OCRE_SHARED_HEAP_BUF_PHYSICAL | ||
| // Physical mode - map hardware register address | ||
| heap_init_args.pre_allocated_addr = (void *)CONFIG_OCRE_SHARED_HEAP_BUF_ADDRESS; |
There was a problem hiding this comment.
This looks like something very dangerous to do as this memory is potentially used by the OS.
There was a problem hiding this comment.
This is to give the user the possibility to manipulate hardware (like GPIOs) from the container - it can be enabled or disabled from the config. But yes, user shall take care and understand, that if they use it it's dangerous.
Update wamr's hash to support networking DNS changes. --------- Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
Update wamr's hash to support networking DNS changes. --------- Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
This reverts commit f555f1b. This commit is kept in the legacy branch.
Update wamr's hash to support networking DNS changes. --------- Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com>
This reverts commit f555f1b. This commit is kept in the legacy branch. Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
Update wamr's hash to support networking DNS changes. --------- Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com> Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
Update wamr's hash to support networking DNS changes. --------- Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com> Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
This reverts commit f555f1b. This commit is kept in the legacy branch. Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
Update wamr's hash to support networking DNS changes. --------- Signed-off-by: Krisztian Szilvasi <34309983+kr-t@users.noreply.github.com> Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
Description
Improvement of wasm stack - additional networking and shared memory support
Type of change
Please delete options that are not relevant.
How Has This Been Tested?
Tests on Linux, Zephyr and CI
Checklist: