Hi, I was recently working on acpi_table_add() changes as a possible workaround for a bug I was chasing. In the end, I dropped the idea of trying to use custom ACPI tables for the specific problem I was working on, but in the process I made some cleanups in the acpi_table_add() function, to make it more readable and easier to change later.
Except for a small bug fix in the error handling (that isn't that important because QEMU exits immediately if acpi_table_add() returns an error), no behavior changes are introduced by this series, being just lots of code movements. Eduardo Habkost (11): acpi_table_add(): simplify acpi_tables on-demand initialization acpi_table_add(): extract acpi_tables init to separate function acpi_table_add(): rename 'len' to 'acpi_len' acpi_table_add(): introduce 'qemu_len' variable acpi_table_add(): extract ACPI header creation to separate function acpi_table_add(): rename 'qemu_len' to 'newlen' acpi_table_add(): eliminate 'allen' variable acpi_table_add(): eliminate 'start' variable acpi_table_add(): extract acpi_tables reallocation code to a separate function acpi_table_add(): use acpi_newtable_resize() return value acpi_table_add(): extract acpi_tables count update code to separate function hw/acpi.c | 185 ++++++++++++++++++++++++++++++++++++------------------------- 1 file changed, 109 insertions(+), 76 deletions(-) -- 1.7.10.4