gint-with-thread/include/gint/gint.h

141 lines
6.3 KiB
C
Raw Permalink Normal View History

//---
// gint - An alternative runtime environment for fx9860g and fxcg50
//---
#ifndef GINT_GINT
#define GINT_GINT
#include <gint/defs/types.h>
/* GINT_VERSION - the library version number
gint is versioned from its repository commits on the master branch. The
GINT_VERSION integer contains the short commit hash with 7 digits.
For instance, 0x03f7c0a0 means commit 3f7c0a0. */
extern char GINT_VERSION;
#define GINT_VERSION ((uint32_t)&GINT_VERSION)
/* gint_switch(): Switch out of gint to execute a function
core, tmu: add gint_switch(), return to menu, and improve timer code * Add the gint_switch() function which executes user-provided code from the system (CASIOWIN) context. * Added interrupt masks to the core context (should have been there long ago). * Added the gint_osmenu() function that switches out of gint to invoke GetKeyWait() and inject KEY_CTRL_MENU to trigger the main menu. This uses many CASIOWIN syscalls, but we don't care because gint is unloaded. Trickery is used to catch the key following the return in the add-in and/or display a new application frame before GetKeyWait() even finishes after coming back. This is only available on fx9860g for now. * Removed any public syscall definition to clear up interfaces. * Patched the DMA interruption problem in a weird way on fxcg50, a driver function will be used to do that properly eventually. * Changed the driver model to save driver contexts in preallocated spaces instead of on the stack for overall less risk. * Enabled return-to-menu with the MENU key on fx9860g in getkey(). * Changed the keyboard driver to emit releases before presses, as a return-to-menu acts as a press+release of different keys in a single driver frame, which confuses getkey(). * Fixed a really stupid bug in memcpy() that made the function really not work. Improvements in the timer driver: * Expose ETMU modules as SH7705_TMU and SH7305_TMU in <gint/mpu/tmu.h>. * Remove the timer_t structures, using SH*_ETMU and SH*_TMU instead. Only interrupt gate entries are left hardcoded. * Discovered that not only every write to the TCNT or TCR of an ETMU takes about 1/32k of a second (hinting at registers being powered by the same clock as the timer), but every write occuring while a previous write is pending is *lost*. This led to terrible bugs when switching ETMU contexts too fast in gint_switch(). * Removed an internal timer_address() function. * Overall simplified the handling of timers and the initialization step.
2020-05-10 14:03:41 +02:00
This function can be used to leave gint, restore the system's driver
context, and execute code there before returning to gint. By doing this one
can effectively interleave gint with the standard OS execution. The
limitations are quite extreme though, so unless you know precisely what
you're doing that requires getting out of gint (eg. BFile), be careful.
core, tmu: add gint_switch(), return to menu, and improve timer code * Add the gint_switch() function which executes user-provided code from the system (CASIOWIN) context. * Added interrupt masks to the core context (should have been there long ago). * Added the gint_osmenu() function that switches out of gint to invoke GetKeyWait() and inject KEY_CTRL_MENU to trigger the main menu. This uses many CASIOWIN syscalls, but we don't care because gint is unloaded. Trickery is used to catch the key following the return in the add-in and/or display a new application frame before GetKeyWait() even finishes after coming back. This is only available on fx9860g for now. * Removed any public syscall definition to clear up interfaces. * Patched the DMA interruption problem in a weird way on fxcg50, a driver function will be used to do that properly eventually. * Changed the driver model to save driver contexts in preallocated spaces instead of on the stack for overall less risk. * Enabled return-to-menu with the MENU key on fx9860g in getkey(). * Changed the keyboard driver to emit releases before presses, as a return-to-menu acts as a press+release of different keys in a single driver frame, which confuses getkey(). * Fixed a really stupid bug in memcpy() that made the function really not work. Improvements in the timer driver: * Expose ETMU modules as SH7705_TMU and SH7305_TMU in <gint/mpu/tmu.h>. * Remove the timer_t structures, using SH*_ETMU and SH*_TMU instead. Only interrupt gate entries are left hardcoded. * Discovered that not only every write to the TCNT or TCR of an ETMU takes about 1/32k of a second (hinting at registers being powered by the same clock as the timer), but every write occuring while a previous write is pending is *lost*. This led to terrible bugs when switching ETMU contexts too fast in gint_switch(). * Removed an internal timer_address() function. * Overall simplified the handling of timers and the initialization step.
2020-05-10 14:03:41 +02:00
This main uses for this switch are going back to the main menu and using
BFile function. You can go back to the main menu easily by calling getkey()
(or getkey_opt() with the GETKEY_MENU flag set) and pressing the MENU key,
or by calling gint_osmenu() below which uses this switch.
core, tmu: add gint_switch(), return to menu, and improve timer code * Add the gint_switch() function which executes user-provided code from the system (CASIOWIN) context. * Added interrupt masks to the core context (should have been there long ago). * Added the gint_osmenu() function that switches out of gint to invoke GetKeyWait() and inject KEY_CTRL_MENU to trigger the main menu. This uses many CASIOWIN syscalls, but we don't care because gint is unloaded. Trickery is used to catch the key following the return in the add-in and/or display a new application frame before GetKeyWait() even finishes after coming back. This is only available on fx9860g for now. * Removed any public syscall definition to clear up interfaces. * Patched the DMA interruption problem in a weird way on fxcg50, a driver function will be used to do that properly eventually. * Changed the driver model to save driver contexts in preallocated spaces instead of on the stack for overall less risk. * Enabled return-to-menu with the MENU key on fx9860g in getkey(). * Changed the keyboard driver to emit releases before presses, as a return-to-menu acts as a press+release of different keys in a single driver frame, which confuses getkey(). * Fixed a really stupid bug in memcpy() that made the function really not work. Improvements in the timer driver: * Expose ETMU modules as SH7705_TMU and SH7305_TMU in <gint/mpu/tmu.h>. * Remove the timer_t structures, using SH*_ETMU and SH*_TMU instead. Only interrupt gate entries are left hardcoded. * Discovered that not only every write to the TCNT or TCR of an ETMU takes about 1/32k of a second (hinting at registers being powered by the same clock as the timer), but every write occuring while a previous write is pending is *lost*. This led to terrible bugs when switching ETMU contexts too fast in gint_switch(). * Removed an internal timer_address() function. * Overall simplified the handling of timers and the initialization step.
2020-05-10 14:03:41 +02:00
@function Function to call in OS mode */
void gint_switch(void (*function)(void));
/* gint_osmenu(): Call the calculator's main menu
This function safely invokes the calculator's main menu with gint_switch().
If the user selects the gint application again in the menu, this function
reloads gint and returns. Otherwise, the add-in is fully unloaded by the
system and the application terminates.
This function is typically called when the [MENU] key is pressed during a
call to getkey(), but can also be called manually. */
core, tmu: add gint_switch(), return to menu, and improve timer code * Add the gint_switch() function which executes user-provided code from the system (CASIOWIN) context. * Added interrupt masks to the core context (should have been there long ago). * Added the gint_osmenu() function that switches out of gint to invoke GetKeyWait() and inject KEY_CTRL_MENU to trigger the main menu. This uses many CASIOWIN syscalls, but we don't care because gint is unloaded. Trickery is used to catch the key following the return in the add-in and/or display a new application frame before GetKeyWait() even finishes after coming back. This is only available on fx9860g for now. * Removed any public syscall definition to clear up interfaces. * Patched the DMA interruption problem in a weird way on fxcg50, a driver function will be used to do that properly eventually. * Changed the driver model to save driver contexts in preallocated spaces instead of on the stack for overall less risk. * Enabled return-to-menu with the MENU key on fx9860g in getkey(). * Changed the keyboard driver to emit releases before presses, as a return-to-menu acts as a press+release of different keys in a single driver frame, which confuses getkey(). * Fixed a really stupid bug in memcpy() that made the function really not work. Improvements in the timer driver: * Expose ETMU modules as SH7705_TMU and SH7305_TMU in <gint/mpu/tmu.h>. * Remove the timer_t structures, using SH*_ETMU and SH*_TMU instead. Only interrupt gate entries are left hardcoded. * Discovered that not only every write to the TCNT or TCR of an ETMU takes about 1/32k of a second (hinting at registers being powered by the same clock as the timer), but every write occuring while a previous write is pending is *lost*. This led to terrible bugs when switching ETMU contexts too fast in gint_switch(). * Removed an internal timer_address() function. * Overall simplified the handling of timers and the initialization step.
2020-05-10 14:03:41 +02:00
void gint_osmenu(void);
/* gint_setrestart(): Set whether to restart the add-in after exiting
An add-in that reaches the end of its code exits. On the calculator, except
using OS-dependent settings, it cannot be started again unless another
application is launched first.
This setting allows the add-in to restart by calling gint_osmenu() instead
of exiting. This can give a proper illusion of restarting if used correctly.
@restart 0 to exit, 1 to restart by using gint_osmenu() */
void gint_setrestart(int restart);
/* gint_inthandler(): Install interrupt handlers
This function installs (copies) interrupt handlers in the VBR space of the
application. Each handler is a 32-byte block aligned on a 32-byte boundary.
When an interrupt request is accepted, the hardware jumps to a specific
interrupt handler at an address that depends on the interrupt source.
For safety, interrupt handlers should avoid referring to data from other
blocks because the arrangement of blocks at runtime depends on event codes.
The assembler program will assume that consecutive blocks in the source code
will be consecutive in memory, which is not always true. Avoiding cross-
references is a practical rule to avoid problems. (gint breaks this rule
quite often but does it safely.)
This function allows anyone to replace any interrupt handler so make sure
you're not interfering with interrupt assignments from gint or a library.
The first parameter event_code represents the event code associated with the
interrupt. If it's not a multiple of 0x20 then you're doing something wrong.
The codes are normally platform-dependent, but gint always uses SH7305
codes. SH3 platforms have a different, compact VBR layout. gint_inthandler()
translates the provided SH7305 codes to the compact layout and the interrupt
handler translates the hardware SH3 codes to the compact layout as well. See
gint's source in <src/core/inth.S> and <src/core/kernel.c>. Please note that
gint_inthandler() uses a table that must be modified for every new SH3
interrupt code to extend the compact scheme.
The handler function is run in the kernel register bank with interrupts
disabled and must end with 'rts' (not 'rte') as the main interrupt handler
saves some registers for you. By default, user bank registers are not saved
except for gbr/mach/macl; if you want to call back to arbitrary code safely,
use gint_inth_callback() in your handler.
For convenience gint allows any block size to be loaded as an interrupt
handler, but it should really be a multiple of 0x20 bytes and not override
other handlers. If it's not written in assembler, then you're likely doing
something wrong. Using __attribute__((interrupt_handler)), which uses rte,
is especially wrong.
It is common for interrupt handlers to have a few bytes of data, such as the
address of a callback function. gint often stores this data in the last
bytes of the block. This function returns the VBR address of the block which
has just been installed, to allow the caller to edit the parameters later.
@event_code Identifier of the interrupt block
@handler Address of handler function
@size How many bytes to copy
Returns the VBR address where the handler was installed. */
void *gint_inthandler(int event_code, void const *handler, size_t size);
/* gint_inth_callback(): Call back arbitrary code from an interrupt handler
Calls the specified function with the given argument after saving the user
context, enabling interrupts and going to user bank. This function is used
to call user code from interrupt handlers, typically from timer or RTC
kernel: dynamic loading of GMAPPED functions to user RAM This commit introduces a large architectural change. Unlike previous models of the fx-9860G series, the G-III models have a new user RAM address different from 8801c000. The purpose of this change is to dynamically load GMAPPED functions to this address by querying the TLB, and call them through a function pointer whose address is determined when loading. Because of the overhead of using a function pointer in both assembly and C code, changes have been made to avoid GMAPPED functions altogether. Current, only cpu_setVBR() and gint_inth_callback() are left, the second being used specifically to enable TLB misses when needed. * Add a .gint.mappedrel section for the function pointers holding addresses to GMAPPED functions; add function pointers for cpu_setVBR() and gint_inth_callback() * Move rram to address 0 instead of the hardcoded 0x8801c000 * Load GMAPPED functions at their linked address + the physical address user RAM is mapped, to and compute their function pointers * Remove the GMAPPED macro since no user function needs it anymore * Add section flags "ax" (code) or "aw" (data) to every custom .section in assembler code, as they default to unpredictable values that can cause the section to be marked NOLOAD by the linker * Update the main kernel, TMU, ETMU and RTC interrupt handlers to use the new indirect calling method This is made possible by new MMU functions giving direct access to the physical area behind any virtualized page. * Add an mmu_translate() function to query the TLB * Add an mmu_uram() function to access user RAM from P1 The exception catching mechanism has been modified to avoid the use of GMAPPED functions altogether. * Set SR.BL=0 and SR.IMASK=15 before calling exception catchers * Move gint_exc_skip() to normal text ROM * Also fix registers not being popped off the stack before a panic The timer drivers have also been modified to avoid GMAPPED functions. * Invoke timer_stop() through gint_inth_callback() and move it to ROM * Move and expand the ETMU driver to span 3 blocks at 0xd00 (ETMU4) * Remove the timer_clear() function by inlining it into the ETMU handler (TCR is provided within the storage block of each timer) * Also split src/timer/inth.s into src/timer/inth-{tmu,etmu}.s Additionally, VBR addresses are now determined at runtime to further reduce hardcoded memory layout addresses in the linker script. * Determine fx-9860G VBR addresses dynamically from mmu_uram() * Determine fx-CG 50 VBR addresses dynamically from mmu_uram() * Remove linker symbols for VBR addresses Comments and documentation have been updated throughout the code to reflect the changes.
2020-09-17 14:48:54 +02:00
callbacks. You can think of it as a way to escape the SR.BL=1 environment to
safely call back virtualized and interrupt-based functions during interrupt
handling.
It is not safe to call from C code in any capacity and is mentioned here
only for documentation purposes; you should really only call this from
kernel: dynamic loading of GMAPPED functions to user RAM This commit introduces a large architectural change. Unlike previous models of the fx-9860G series, the G-III models have a new user RAM address different from 8801c000. The purpose of this change is to dynamically load GMAPPED functions to this address by querying the TLB, and call them through a function pointer whose address is determined when loading. Because of the overhead of using a function pointer in both assembly and C code, changes have been made to avoid GMAPPED functions altogether. Current, only cpu_setVBR() and gint_inth_callback() are left, the second being used specifically to enable TLB misses when needed. * Add a .gint.mappedrel section for the function pointers holding addresses to GMAPPED functions; add function pointers for cpu_setVBR() and gint_inth_callback() * Move rram to address 0 instead of the hardcoded 0x8801c000 * Load GMAPPED functions at their linked address + the physical address user RAM is mapped, to and compute their function pointers * Remove the GMAPPED macro since no user function needs it anymore * Add section flags "ax" (code) or "aw" (data) to every custom .section in assembler code, as they default to unpredictable values that can cause the section to be marked NOLOAD by the linker * Update the main kernel, TMU, ETMU and RTC interrupt handlers to use the new indirect calling method This is made possible by new MMU functions giving direct access to the physical area behind any virtualized page. * Add an mmu_translate() function to query the TLB * Add an mmu_uram() function to access user RAM from P1 The exception catching mechanism has been modified to avoid the use of GMAPPED functions altogether. * Set SR.BL=0 and SR.IMASK=15 before calling exception catchers * Move gint_exc_skip() to normal text ROM * Also fix registers not being popped off the stack before a panic The timer drivers have also been modified to avoid GMAPPED functions. * Invoke timer_stop() through gint_inth_callback() and move it to ROM * Move and expand the ETMU driver to span 3 blocks at 0xd00 (ETMU4) * Remove the timer_clear() function by inlining it into the ETMU handler (TCR is provided within the storage block of each timer) * Also split src/timer/inth.s into src/timer/inth-{tmu,etmu}.s Additionally, VBR addresses are now determined at runtime to further reduce hardcoded memory layout addresses in the linker script. * Determine fx-9860G VBR addresses dynamically from mmu_uram() * Determine fx-CG 50 VBR addresses dynamically from mmu_uram() * Remove linker symbols for VBR addresses Comments and documentation have been updated throughout the code to reflect the changes.
2020-09-17 14:48:54 +02:00
an interrupt handler's assembler code, typically like this:
mov.l .callback, r0
mov.l @r0, r0 # because function pointer
mov <function>, r4
mov <argument>, r5
jsr @r0
nop
.callback:
.long _gint_inth_callback
This function is loaded to a platform-dependent address determined at
runtime; call it indirectly through the function pointer.
@callback Callback function, may take no argument in practice
@arg Argument
Returns the return value of the callback. */
extern int (*gint_inth_callback)(int (*function)(void *arg), void *arg);
#endif /* GINT_GINT */