

@189

5 years 
max 
sync



@187

5 years 
max 
count the number of IOAPICs



@169

5 years 
max 
remove unused var



@168

5 years 
max 
make the TLS initialization percpu



@167

5 years 
max 
define the TLS in hal_segmentation.h



@166

5 years 
max 
identify the cpu features



@165

5 years 
max 
put the percpu structures into one structure



@164

5 years 
max 
we're gonna use an array of TSSs



@163

5 years 
max 
handle the LAPIC override, and count the number of CPUs



@162

5 years 
max 
start making these variables percpu



@155

5 years 
max 
cleanup



@154

5 years 
max 
use a demultiplexer to get the channel on the serial port, and
make …



@153

5 years 
max 
cleanup



@152

5 years 
max 
add a basic RS232 COM1 implementation



@150

5 years 
max 
remove debug and unused code



@148

5 years 
max 
fix a pretty dumb bug; we must not unmap since we use the area for …



@147

5 years 
max 
less magic



@146

5 years 
max 
remove my fake_apic shit from yesterday



@145

5 years 
max 
style



@144

5 years 
max 
remove unused files



@143

5 years 
max 
style



@142

5 years 
max 
remove hal_trap.c, and put its content in hal_interrupt.c and …



@139

5 years 
max 
I'm dumb



@138

5 years 
max 
update



@137

5 years 
max 
improve the APIC implementation



@136

5 years 
max 
Hide a few soclib_xcu_ functions



@135

5 years 
max 
start moving the APIC into the XCU driver



@134

5 years 
max 
fix comment



@130

5 years 
max 
add hal_drivers_xcu_init



@129

5 years 
max 
put panics everywhere



@125

5 years 
max 
implement hal_time_stamp and hal_fence



@124

5 years 
max 
rename hal_wbflush>hal_fence



@123

5 years 
max 
unmap the area below the kernel, so that we can find pointer …



@122

5 years 
max 
we will fix nested interrupt contexts later



@119

5 years 
max 
provide pages_nr



@118

5 years 
max 
add hal_restore_irq



@117

5 years 
max 
Calibrate the LAPIC timer with the PIT, with one interrupt per
second …



@116

5 years 
max 
complete hal_ppm_init, to take into account the reserved areas



@115

5 years 
max 
remove duplicate



@114

5 years 
max 
compute the offset for bootinfo



@113

5 years 
max 
add memcmp in libk; these mem* functions should probably be hal …



@112

5 years 
max 
In fact, we need to enter all of the VA space, because we massively …



@111

5 years 
max 
Mmh, update CLUSTER_VA_SIZE, we are calibrated for 64GB



@110

5 years 
max 
add some code for ppm_init



@109

5 years 
max 
don't save/restore %gs for now, we want to use TLS in interrupt …



@105

5 years 
max 
add a few atomic functions



@102

5 years 
max 
update the new hal_get_cycles



@101

5 years 
alain 
euh…



@100

5 years 
max 
add a part of the tty driver; not tested yet (due to a NULL deref …



@99

5 years 
max 
add the irq functions; we will forbid nested critical sections, so
the …



@98

5 years 
max 
add the and/or ops



@97

5 years 
max 
(forgot this one in my previous commit)



@96

5 years 
max 
Mmh, fix the return values



@94

5 years 
max 
remove lw_unc, add a few ops, and update a few things



@92

5 years 
max 
fix the definition of hal_remote_spt, and add the *pt ops on
x86_64



@91

5 years 
max 
retrieve the pressed key, and display it



@90

5 years 
max 
scroll for real, and add %c



@89

5 years 
max 
Parse the IOAPIC structure in ACPI, map the IOAPIC, and enable
the …



@86

5 years 
max 
enable the timer (not yet calibrated)



@85

5 years 
max 
reload the segregs



@84

5 years 
max 
scroll the console



@83

5 years 
max 
we want to have the LAPIC noncacheable



@82

5 years 
max 
Rename a certain number of things, and improve the APIC support.



@81

5 years 
max 
silence a few warnings



@80

5 years 
max 
add the timer vector, for now it doublefaults



@79

5 years 
max 
invalidate the page directly



@78

5 years 
max 
print %rip, and add cli/sti



@77

5 years 
max 
Start hiding the architecturespecific drivers behind the …



@76

5 years 
max 
empty out the x86 drivers



@75

5 years 
max 
Create the drivers/ subdirectory in each hal, and move soclib
into …



@74

5 years 
max 
(forgot to commit this)



@73

5 years 
max 
now that the fields are stable, embed the VA base in XPTR



@72

5 years 
max 
start defining some remote functions, and use uint8_t instead of
char



@71

5 years 
max 
advance a little



@70

5 years 
max 
start constructing the bootinfo structure



@69

5 years 
max 
use uint8_t instead



@67

5 years 
max 
switch to 64GB of VA per cluster, with 256 clusters



@61

5 years 
max 
cosmetic



@57

5 years 
max 
move the ldscript into x86_64/



@54

5 years 
max 
make it compile on x86



@51

5 years 
max 
Create the core/ subdirectory for x86_64.



@48

5 years 
max 
Put panics everywhere



@47

5 years 
max 
Use mcmodel=large, in order to have a kernel image that is located at …



@46

5 years 
max 
Start implementing TLS.



@45

5 years 
max 
Add some code for LAPIC; far from complete, but a good start.



@44

5 years 
max 
Update. We need to use two separate arrays: one for the heap and one …



@43

5 years 
max 
simplify



@42

5 years 
max 
Explicitly use configurable parameters



@39

5 years 
max 
Parse RSDP>RSDT>MADT, and get the LAPIC PA.



@38

5 years 
max 
Use 4GB of VA per cluster, and 8GB of PA.



@35

5 years 
max 
Update:
* Introduce a basic VM manager in hal_gpt.c, to create the …



@34

5 years 
max 
Update the types (from i386/).



@32

5 years 
max 
Update. Retrieve and parse the multiboot info, and dump the mmap. Some …



@31

5 years 
max 
Silence a few gcc warnings.



@29

5 years 
max 
Update. The kernel now enables the GDT/IDT, and has trap entries. A …



@25

5 years 
max 
First shot of the x86_64 port. It uses Multiboot V1. Most of the HAL …
