Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
Donny9 commented on PR #12188: URL: https://github.com/apache/nuttx/pull/12188#issuecomment-2068590626 > > > @Donny9 in rwb_wrflush(): memmove(rwb->wrbuffer + padblocks * rwb->blocksize, rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize); > > > when the offest from rwb->wrblockstart to startblock plus rwb->wrnblocks is greater than rwb->wralignblocks, it will be causing memory overflow since rwb->wrbuffer is allocated rwb.wrmaxblock(rwb.wralignblocks) blocks. > > > For more detail，we can communicate in wechat:chen_chuang_1030 > > > > > > @ChenChuangForLinux i got, I will improve a new patch to solve this problem instead of expanding the buffer > > OK, maybe rwb_writebuffer() is a good starting point @ChenChuangForLinux you can try this patch https://github.com/apache/nuttx/pull/12202 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux commented on PR #12188: URL: https://github.com/apache/nuttx/pull/12188#issuecomment-2067965309 > > @Donny9 in rwb_wrflush(): memmove(rwb->wrbuffer + padblocks * rwb->blocksize, rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize); > > when the offest from rwb->wrblockstart to startblock plus rwb->wrnblocks is greater than rwb->wralignblocks, it will be causing memory overflow since rwb->wrbuffer is allocated rwb.wrmaxblock(rwb.wralignblocks) blocks. > > For more detail，we can communicate in wechat:chen_chuang_1030 > > @ChenChuangForLinux i got, I will improve a new patch to solve this problem instead of expanding the buffer OK, maybe rwb_writebuffer() is a good starting point -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
Donny9 commented on PR #12188: URL: https://github.com/apache/nuttx/pull/12188#issuecomment-2067962532 > @Donny9 in rwb_wrflush(): memmove(rwb->wrbuffer + padblocks * rwb->blocksize, rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize); > > when the offest from rwb->wrblockstart to startblock plus rwb->wrnblocks is greater than rwb->wralignblocks, it will be causing memory overflow since rwb->wrbuffer is allocated rwb.wrmaxblock(rwb.wralignblocks) blocks. > > For more detail，we can communicate in wechat:chen_chuang_1030 @ChenChuangForLinux i got, I will improve a new patch to solve this problem instead of expanding the buffer -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux commented on PR #12188: URL: https://github.com/apache/nuttx/pull/12188#issuecomment-2067942471 @Donny9 in rwb_wrflush(): memmove(rwb->wrbuffer + padblocks * rwb->blocksize, rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize); when the offest from rwb->wrblockstart to startblock plus rwb->wrnblocks is greater than rwb->wralignblocks, it will be causing memory overflow since rwb->wrbuffer is allocated rwb.wrmaxblock(rwb.wralignblocks) blocks. For more detail，we can communicate in wechat:chen_chuang_1030 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
Donny9 commented on PR #12188: URL: https://github.com/apache/nuttx/pull/12188#issuecomment-2067937759 @ChenChuangForLinux In what specific function will this overflow occur? Why can't we write in two parts? First align the first part and flush it, and then put the second part into the wrbuffer. The writebuffer is already large for nand flsh with 256K eraseblock, double it will be 512K -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
acassis merged PR #12188: URL: https://github.com/apache/nuttx/pull/12188 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
[PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux opened a new pull request, #12188: URL: https://github.com/apache/nuttx/pull/12188 when the offest from rwb->wrblockstart to startblock plus rwb->wrnblocks is greater than rwb->wralignblocks, it will be causing memory overflow since rwb->wrbuffer is allocated rwb.wrmaxblock(rwb.wralignblocks) blocks. ## Summary ## Impact ## Testing -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux closed pull request #12056: fix: the bug about rwbuffer causing mem overflow URL: https://github.com/apache/nuttx/pull/12056 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
xiaoxiang781216 commented on PR #12056: URL: https://github.com/apache/nuttx/pull/12056#issuecomment-2053981191 @ChenChuangForLinux please remove the merge patch from pr. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
xiaoxiang781216 commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1564585552
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,50 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ uint8_t *tmp = NULL;
+ size_t tmp_blocks = 0;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ if (padblocks + rwb->wrnblocks > rwb->wralignblocks)
+{
+ tmp_blocks = padblocks - rwb->wrnblocks -
+rwb->wralignblocks;
+ tmp = kmm_malloc(tmp_blocks * rwb->blocksize);
Review Comment:
it isn't good to allocate temp buffer in the hot path.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1564556600
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
The latest PR is my final solution based on the original idea.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1564556600
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
This my final solution based on the original idea.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
chenchuangchn commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1555010494
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
But in another case in rwb_writebuffer()(rwbuffer.c:367), startblock may be
unalignment, it also just call rwb->wrflush(ftl_flush).
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
xiaoxiang781216 commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554853308
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
No, ftl depends on the rwb alignment at all.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554839997
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
Sorry, I'm busy today. Why not remove all unalignment issue handler in
rwb_wrflush()? ftl layler has been done it.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
xiaoxiang781216 commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554834154
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
After more thinking, it's hard to handle unalignment issue here when the
cache is near full without additional buffer. So, I think the better approach
may allocate the buffer with one additional aligned block, or ensure the
startblock is aligned at the first place.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554798912
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
Oh, you're right, it my fault. I will take a new PR in 3 hours. By the way,
I don't think it't necessarily to use padblocks strategy in rwb_wrflush(), just
call rwb->wrflush() is simple, since ftl layer has done enough precoess.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
xiaoxiang781216 commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554656121
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
Since wrnblocks may smaller than rwb->wrnblocks, [rwb->wrnblocks -
wrnblocks, rwb->wrnblocks) is destroyed by memmove.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
chenchuangchn commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554632078
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
rwb->wrblockstart -= padblocks;
- rwb->wrnblocks += padblocks;
+ rwb->wrnblocks -= wrnblocks;
rwb_read_(rwb, rwb->wrblockstart, padblocks, rwb->wrbuffer);
+
+ ret = rwb->wrflush(rwb->dev, rwb->wrbuffer, rwb->wrblockstart,
+ padblocks + wrnblocks);
Review Comment:
OK
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
chenchuangchn commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554631996
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
The padblocks + wrnblocks <= wralignblocks, why will the memmove destroy the
content at the end?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
acassis commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554621902
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
rwb->wrblockstart -= padblocks;
- rwb->wrnblocks += padblocks;
+ rwb->wrnblocks -= wrnblocks;
rwb_read_(rwb, rwb->wrblockstart, padblocks, rwb->wrbuffer);
+
+ ret = rwb->wrflush(rwb->dev, rwb->wrbuffer, rwb->wrblockstart,
+ padblocks + wrnblocks);
Review Comment:
```suggestion
padblocks + wrnblocks);
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
xiaoxiang781216 commented on code in PR #12056:
URL: https://github.com/apache/nuttx/pull/12056#discussion_r1554582306
##
drivers/misc/rwbuffer.c:
##
@@ -140,18 +140,28 @@ static void rwb_wrflush(FAR struct rwbuffer_s *rwb)
if (rwb->wrnblocks > 0)
{
size_t padblocks;
+ size_t wrnblocks;
finfo("Flushing: blockstart=0x%08lx nblocks=%d from buffer=%p\n",
(long)rwb->wrblockstart, rwb->wrnblocks, rwb->wrbuffer);
padblocks = rwb->wrblockstart % rwb->wralignblocks;
if (padblocks)
{
+ wrnblocks = (rwb->wralignblocks - padblocks) < rwb->wrnblocks ?
+(rwb->wralignblocks - padblocks) : rwb->wrnblocks;
memmove(rwb->wrbuffer + padblocks * rwb->blocksize,
- rwb->wrbuffer, rwb->wrnblocks * rwb->blocksize);
+ rwb->wrbuffer, wrnblocks * rwb->blocksize);
Review Comment:
but, memmove with partial block will destroy the content at the end,
@ChenChuangForLinux .
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
xiaoxiang781216 commented on PR #12056: URL: https://github.com/apache/nuttx/pull/12056#issuecomment-2040970781 @ChenChuangForLinux please squash into one patch. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
acassis commented on PR #12056: URL: https://github.com/apache/nuttx/pull/12056#issuecomment-2039747255 @ChenChuangForLinux please check coding style issues: $ ./tools/checkpatch.sh -f drivers/misc/rwbuffer.c And fix the issues -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
Re: [PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux commented on PR #12056: URL: https://github.com/apache/nuttx/pull/12056#issuecomment-2039013658 when the offest from rwb->wrblockstart to startblock plus rwb->wrnblocks is greater than rwb->wralignblocks, it will be causing memory overflow since rwb->wrbuffer is allocated rwb.wrmaxblock(rwb.wralignblocks) block_size bytes -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
[PR] fix: the bug about rwbuffer causing mem overflow [nuttx]
ChenChuangForLinux opened a new pull request, #12056: URL: https://github.com/apache/nuttx/pull/12056 when the offest from rwb->wrblockstart to startblock plus rwb->wrnblocks is greater than rwb->wralignblocks, it will be causing memory overflow since rwb->wrbuffer is allocated rwb.wrmaxblock(rwb.wralignblocks) bytes. ## Summary ## Impact ## Testing -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@nuttx.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
