From: Vincent Chen <vincent.chen@sifive.com>
To: paul.walmsley@sifive.com, palmer@sifive.com, aou@eecs.berkeley.edu
Cc: Vincent Chen <vincent.chen@sifive.com>,
linux-riscv@lists.infradead.org, linux-kernel@vger.kernel.org
Subject: [PATCH 1/2] riscv: Correct the initialized flow of FP register
Date: Thu, 8 Aug 2019 15:58:40 +0800 [thread overview]
Message-ID: <1565251121-28490-2-git-send-email-vincent.chen@sifive.com> (raw)
In-Reply-To: <1565251121-28490-1-git-send-email-vincent.chen@sifive.com>
The following two reasons cause FP registers are sometimes not
initialized before starting the user program.
1. Currently, the FP context is initialized in flush_thread() function
and we expect these initial values to be restored to FP register when
doing FP context switch. However, the FP context switch only occurs in
switch_to function. Hence, if this process does not be scheduled out
and scheduled in before entering the user space, the FP registers
have no chance to initialize.
2. In flush_thread(), the state of reg->sstatus.FS inherits from the
parent. Hence, the state of reg->sstatus.FS may be dirty. If this
process is scheduled out during flush_thread() and initializing the
FP register, the fstate_save() in switch_to will corrupt the FP context
which has been initialized until flush_thread().
To solve the 1st case, the initialization of the FP register will be
completed in start_thread(). It makes sure all FP registers are initialized
before starting the user program. For the 2nd case, the state of
reg->sstatus.FS in start_thread will be set to SR_FS_OFF to prevent this
process from corrupting FP context in doing context save. The FP state is
set to SR_FS_INITIAL in start_trhead().
Tested on both QEMU and HiFive Unleashed using BBL + Linux.
Signed-off-by: Vincent Chen <vincent.chen@sifive.com>
---
arch/riscv/include/asm/switch_to.h | 6 ++++++
arch/riscv/kernel/process.c | 13 +++++++++++--
2 files changed, 17 insertions(+), 2 deletions(-)
diff --git a/arch/riscv/include/asm/switch_to.h b/arch/riscv/include/asm/switch_to.h
index 853b65e..d5fe573 100644
--- a/arch/riscv/include/asm/switch_to.h
+++ b/arch/riscv/include/asm/switch_to.h
@@ -19,6 +19,12 @@ static inline void __fstate_clean(struct pt_regs *regs)
regs->sstatus |= (regs->sstatus & ~(SR_FS)) | SR_FS_CLEAN;
}
+static inline void fstate_off(struct task_struct *task,
+ struct pt_regs *regs)
+{
+ regs->sstatus = (regs->sstatus & ~(SR_FS)) | SR_FS_OFF;
+}
+
static inline void fstate_save(struct task_struct *task,
struct pt_regs *regs)
{
diff --git a/arch/riscv/kernel/process.c b/arch/riscv/kernel/process.c
index f23794b..e3077ee 100644
--- a/arch/riscv/kernel/process.c
+++ b/arch/riscv/kernel/process.c
@@ -64,8 +64,16 @@ void start_thread(struct pt_regs *regs, unsigned long pc,
unsigned long sp)
{
regs->sstatus = SR_SPIE;
- if (has_fpu)
+ if (has_fpu) {
regs->sstatus |= SR_FS_INITIAL;
+#ifdef CONFIG_FPU
+ /*
+ * Restore the initial value to the FP register
+ * before starting the user program.
+ */
+ fstate_restore(current, regs);
+#endif
+ }
regs->sepc = pc;
regs->sp = sp;
set_fs(USER_DS);
@@ -75,10 +83,11 @@ void flush_thread(void)
{
#ifdef CONFIG_FPU
/*
- * Reset FPU context
+ * Reset FPU state and context
* frm: round to nearest, ties to even (IEEE default)
* fflags: accrued exceptions cleared
*/
+ fstate_off(current, task_pt_regs(current));
memset(¤t->thread.fstate, 0, sizeof(current->thread.fstate));
#endif
}
--
2.7.4
_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv
next prev parent reply other threads:[~2019-08-08 8:00 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-08-08 7:58 [PATCH 0/2] riscv: Correct the initialized flow of FP and __fstate_clean() Vincent Chen
2019-08-08 7:58 ` Vincent Chen [this message]
2019-08-08 10:15 ` [PATCH 1/2] riscv: Correct the initialized flow of FP register Anup Patel
2019-08-08 15:50 ` Paul Walmsley
2019-08-14 1:45 ` Vincent Chen
2019-08-14 3:51 ` Anup Patel
2019-08-12 14:58 ` Christoph Hellwig
2019-08-14 1:52 ` Vincent Chen
2019-08-08 7:58 ` [PATCH 2/2] riscv: Make __fstate_clean() can work correctly Vincent Chen
2019-08-08 10:17 ` Anup Patel
2019-08-14 1:55 ` Vincent Chen
2019-08-12 14:59 ` Christoph Hellwig
2019-08-14 1:55 ` Vincent Chen
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=1565251121-28490-2-git-send-email-vincent.chen@sifive.com \
--to=vincent.chen@sifive.com \
--cc=aou@eecs.berkeley.edu \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-riscv@lists.infradead.org \
--cc=palmer@sifive.com \
--cc=paul.walmsley@sifive.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).