All of lore.kernel.org
 help / color / mirror / Atom feed
From: Alex Ghiti <alex@ghiti.fr>
To: Vitaly Wool <vitaly.wool@konsulko.com>, linux-riscv@lists.infradead.org
Cc: linux-kernel@vger.kernel.org,
	Palmer Dabbelt <palmerdabbelt@google.com>,
	Nicolas Pitre <nico@fluxnic.net>
Subject: Re: [PATCH] riscv: xip: support runtime trap patching
Date: Fri, 4 Jun 2021 11:35:17 +0200	[thread overview]
Message-ID: <2fb64dce-779a-c6f2-53b7-088cd38813a4@ghiti.fr> (raw)
In-Reply-To: <62961217-1733-eaa3-b072-cbb252620c9a@ghiti.fr>

Le 31/05/2021 à 17:17, Alex Ghiti a écrit :
> Hi Vitaly,
> 
> Le 31/05/2021 à 10:53, Vitaly Wool a écrit :
>> RISCV_ERRATA_ALTERNATIVE patches text at runtime which is currently
>> not possible when the kernel is executed from the flash in XIP mode.
>> Since runtime patching concerns only traps at the moment, let's just
>> have all the traps reside in RAM anyway if RISCV_ERRATA_ALTERNATIVE
>> is set. Thus, these functions will be patch-able even when the .text
>> section is in flash.
>>
> 
> This sounds like a good fix for sifive platforms to work with XIP kernel 
> in 5.13: did you test that it actually works on HW?
> 
>> Signed-off-by: Vitaly Wool <vitaly.wool@konsulko.com>
>> ---
>>   arch/riscv/kernel/traps.c           | 13 +++++++++----
>>   arch/riscv/kernel/vmlinux-xip.lds.S | 15 ++++++++++++++-
>>   2 files changed, 23 insertions(+), 5 deletions(-)
>>
>> diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
>> index 0721b9798595..7bc88d8aab97 100644
>> --- a/arch/riscv/kernel/traps.c
>> +++ b/arch/riscv/kernel/traps.c
>> @@ -86,8 +86,13 @@ static void do_trap_error(struct pt_regs *regs, int 
>> signo, int code,
>>       }
>>   }
>> +#if defined (CONFIG_XIP_KERNEL) && defined 
>> (CONFIG_RISCV_ERRATA_ALTERNATIVE)
>> +#define __trap_section        __section(".xip.traps")
>> +#else
>> +#define __trap_section
>> +#endif
> 
> Maybe we could do something more generic. At the moment, only traps are 
> subject to alternatives but that will likely expand: what about rather 
> defining a section called __alternative_section?

Any thoughts about that?

Thanks,

Alex

> 
>>   #define DO_ERROR_INFO(name, signo, code, str)                \
>> -asmlinkage __visible void name(struct pt_regs *regs)            \
>> +asmlinkage __visible __trap_section void name(struct pt_regs *regs)    \
>>   {                                    \
>>       do_trap_error(regs, signo, code, regs->epc, "Oops - " str);    \
>>   }
>> @@ -111,7 +116,7 @@ DO_ERROR_INFO(do_trap_store_misaligned,
>>   int handle_misaligned_load(struct pt_regs *regs);
>>   int handle_misaligned_store(struct pt_regs *regs);
>> -asmlinkage void do_trap_load_misaligned(struct pt_regs *regs)
>> +asmlinkage void __trap_section do_trap_load_misaligned(struct pt_regs 
>> *regs)
>>   {
>>       if (!handle_misaligned_load(regs))
>>           return;
>> @@ -119,7 +124,7 @@ asmlinkage void do_trap_load_misaligned(struct 
>> pt_regs *regs)
>>                 "Oops - load address misaligned");
>>   }
>> -asmlinkage void do_trap_store_misaligned(struct pt_regs *regs)
>> +asmlinkage void __trap_section do_trap_store_misaligned(struct 
>> pt_regs *regs)
>>   {
>>       if (!handle_misaligned_store(regs))
>>           return;
>> @@ -146,7 +151,7 @@ static inline unsigned long 
>> get_break_insn_length(unsigned long pc)
>>       return GET_INSN_LENGTH(insn);
>>   }
>> -asmlinkage __visible void do_trap_break(struct pt_regs *regs)
>> +asmlinkage __visible __trap_section void do_trap_break(struct pt_regs 
>> *regs)
>>   {
>>   #ifdef CONFIG_KPROBES
>>       if (kprobe_single_step_handler(regs))
>> diff --git a/arch/riscv/kernel/vmlinux-xip.lds.S 
>> b/arch/riscv/kernel/vmlinux-xip.lds.S
>> index 4b29b9917f99..a3ff09c4c3f9 100644
>> --- a/arch/riscv/kernel/vmlinux-xip.lds.S
>> +++ b/arch/riscv/kernel/vmlinux-xip.lds.S
>> @@ -99,9 +99,22 @@ SECTIONS
>>       }
>>       PERCPU_SECTION(L1_CACHE_BYTES)
>> -    . = ALIGN(PAGE_SIZE);
>> +    . = ALIGN(8);
>> +    .alternative : {
>> +        __alt_start = .;
>> +        *(.alternative)
>> +        __alt_end = .;
>> +    }
>>       __init_end = .;
>> +    . = ALIGN(16);
> 
> Why 16 here?
> 
>> +    .xip.traps : {
>> +        __xip_traps_start = .;
>> +        *(.xip.traps)
>> +        __xip_traps_end = .;
>> +    }
>> +
>> +    . = ALIGN(PAGE_SIZE);
>>       .sdata : {
>>           __global_pointer$ = . + 0x800;
>>           *(.sdata*)
>>
> 
> Thanks,
> 
> Alex
> 
> _______________________________________________
> linux-riscv mailing list
> linux-riscv@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-riscv

WARNING: multiple messages have this Message-ID (diff)
From: Alex Ghiti <alex@ghiti.fr>
To: Vitaly Wool <vitaly.wool@konsulko.com>, linux-riscv@lists.infradead.org
Cc: linux-kernel@vger.kernel.org,
	Palmer Dabbelt <palmerdabbelt@google.com>,
	Nicolas Pitre <nico@fluxnic.net>
Subject: Re: [PATCH] riscv: xip: support runtime trap patching
Date: Fri, 4 Jun 2021 11:35:17 +0200	[thread overview]
Message-ID: <2fb64dce-779a-c6f2-53b7-088cd38813a4@ghiti.fr> (raw)
In-Reply-To: <62961217-1733-eaa3-b072-cbb252620c9a@ghiti.fr>

Le 31/05/2021 à 17:17, Alex Ghiti a écrit :
> Hi Vitaly,
> 
> Le 31/05/2021 à 10:53, Vitaly Wool a écrit :
>> RISCV_ERRATA_ALTERNATIVE patches text at runtime which is currently
>> not possible when the kernel is executed from the flash in XIP mode.
>> Since runtime patching concerns only traps at the moment, let's just
>> have all the traps reside in RAM anyway if RISCV_ERRATA_ALTERNATIVE
>> is set. Thus, these functions will be patch-able even when the .text
>> section is in flash.
>>
> 
> This sounds like a good fix for sifive platforms to work with XIP kernel 
> in 5.13: did you test that it actually works on HW?
> 
>> Signed-off-by: Vitaly Wool <vitaly.wool@konsulko.com>
>> ---
>>   arch/riscv/kernel/traps.c           | 13 +++++++++----
>>   arch/riscv/kernel/vmlinux-xip.lds.S | 15 ++++++++++++++-
>>   2 files changed, 23 insertions(+), 5 deletions(-)
>>
>> diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
>> index 0721b9798595..7bc88d8aab97 100644
>> --- a/arch/riscv/kernel/traps.c
>> +++ b/arch/riscv/kernel/traps.c
>> @@ -86,8 +86,13 @@ static void do_trap_error(struct pt_regs *regs, int 
>> signo, int code,
>>       }
>>   }
>> +#if defined (CONFIG_XIP_KERNEL) && defined 
>> (CONFIG_RISCV_ERRATA_ALTERNATIVE)
>> +#define __trap_section        __section(".xip.traps")
>> +#else
>> +#define __trap_section
>> +#endif
> 
> Maybe we could do something more generic. At the moment, only traps are 
> subject to alternatives but that will likely expand: what about rather 
> defining a section called __alternative_section?

Any thoughts about that?

Thanks,

Alex

> 
>>   #define DO_ERROR_INFO(name, signo, code, str)                \
>> -asmlinkage __visible void name(struct pt_regs *regs)            \
>> +asmlinkage __visible __trap_section void name(struct pt_regs *regs)    \
>>   {                                    \
>>       do_trap_error(regs, signo, code, regs->epc, "Oops - " str);    \
>>   }
>> @@ -111,7 +116,7 @@ DO_ERROR_INFO(do_trap_store_misaligned,
>>   int handle_misaligned_load(struct pt_regs *regs);
>>   int handle_misaligned_store(struct pt_regs *regs);
>> -asmlinkage void do_trap_load_misaligned(struct pt_regs *regs)
>> +asmlinkage void __trap_section do_trap_load_misaligned(struct pt_regs 
>> *regs)
>>   {
>>       if (!handle_misaligned_load(regs))
>>           return;
>> @@ -119,7 +124,7 @@ asmlinkage void do_trap_load_misaligned(struct 
>> pt_regs *regs)
>>                 "Oops - load address misaligned");
>>   }
>> -asmlinkage void do_trap_store_misaligned(struct pt_regs *regs)
>> +asmlinkage void __trap_section do_trap_store_misaligned(struct 
>> pt_regs *regs)
>>   {
>>       if (!handle_misaligned_store(regs))
>>           return;
>> @@ -146,7 +151,7 @@ static inline unsigned long 
>> get_break_insn_length(unsigned long pc)
>>       return GET_INSN_LENGTH(insn);
>>   }
>> -asmlinkage __visible void do_trap_break(struct pt_regs *regs)
>> +asmlinkage __visible __trap_section void do_trap_break(struct pt_regs 
>> *regs)
>>   {
>>   #ifdef CONFIG_KPROBES
>>       if (kprobe_single_step_handler(regs))
>> diff --git a/arch/riscv/kernel/vmlinux-xip.lds.S 
>> b/arch/riscv/kernel/vmlinux-xip.lds.S
>> index 4b29b9917f99..a3ff09c4c3f9 100644
>> --- a/arch/riscv/kernel/vmlinux-xip.lds.S
>> +++ b/arch/riscv/kernel/vmlinux-xip.lds.S
>> @@ -99,9 +99,22 @@ SECTIONS
>>       }
>>       PERCPU_SECTION(L1_CACHE_BYTES)
>> -    . = ALIGN(PAGE_SIZE);
>> +    . = ALIGN(8);
>> +    .alternative : {
>> +        __alt_start = .;
>> +        *(.alternative)
>> +        __alt_end = .;
>> +    }
>>       __init_end = .;
>> +    . = ALIGN(16);
> 
> Why 16 here?
> 
>> +    .xip.traps : {
>> +        __xip_traps_start = .;
>> +        *(.xip.traps)
>> +        __xip_traps_end = .;
>> +    }
>> +
>> +    . = ALIGN(PAGE_SIZE);
>>       .sdata : {
>>           __global_pointer$ = . + 0x800;
>>           *(.sdata*)
>>
> 
> Thanks,
> 
> Alex
> 
> _______________________________________________
> linux-riscv mailing list
> linux-riscv@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-riscv

_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv

  parent reply	other threads:[~2021-06-04  9:35 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-31  8:53 [PATCH] riscv: xip: support runtime trap patching Vitaly Wool
2021-05-31  8:53 ` Vitaly Wool
2021-05-31 15:17 ` Alex Ghiti
2021-05-31 15:17   ` Alex Ghiti
2021-06-01 10:48   ` Vitaly Wool
2021-06-01 10:48     ` Vitaly Wool
2021-06-04  9:35   ` Alex Ghiti [this message]
2021-06-04  9:35     ` Alex Ghiti
2021-06-04  9:52     ` Vitaly Wool
2021-06-04  9:52       ` Vitaly Wool
2021-06-11  0:51       ` Palmer Dabbelt
2021-06-11  0:51         ` Palmer Dabbelt

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=2fb64dce-779a-c6f2-53b7-088cd38813a4@ghiti.fr \
    --to=alex@ghiti.fr \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-riscv@lists.infradead.org \
    --cc=nico@fluxnic.net \
    --cc=palmerdabbelt@google.com \
    --cc=vitaly.wool@konsulko.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.